Thermochronological, petrographic and geochemical characteristics of the Combia Formation, Amagá basin, Colombia Matthias Bernet, Juliana Mesa Garcia, Catherine Chauvel, Maria Ramírez Londoño, Maria Marín-Cerón

To cite this version:

Matthias Bernet, Juliana Mesa Garcia, Catherine Chauvel, Maria Ramírez Londoño, Maria Marín- Cerón. Thermochronological, petrographic and geochemical characteristics of the Combia Forma- tion, Amagá basin, Colombia. Journal of South American Earth Sciences, Elsevier, 2020, 104, ￿10.1016/j.jsames.2020.102897￿. ￿hal-02990433￿

HAL Id: hal-02990433 https://hal.archives-ouvertes.fr/hal-02990433 Submitted on 17 Nov 2020

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. 1 Thermochronological, petrographic and geochemical 2 characteristics of the Combia Formation, Amagá basin, Colombia

3

4 Matthias Bernet1*, Juliana Mesa Garcia2,3, Catherine Chauvel1,4

5 and Maria Isabel Marín-Cerón2,

6 1Institut des Sciences de la Terre, CNRS, Université Grenoble Alpes, Grenoble, 7 France 8 2Departemento de Geociencias, Universidad EAFIT, Medellín, Colombia 9 10 3present address: Geology Department, University of Michigan, Ann Arbor, MI, USA 11 12 4Université de Paris, Institut de Physique du Globe de Paris, CNRS,F-75005 Paris, 13 France 14 15 *corresponding author, email: [email protected], 16 ORCID: 0000-0001-5046-7520 17

18

19 Abstract

20 The Amagá basin between the Western and Central Cordilleras of the

21 Northern Andes of Colombia host the Neogene volcanic and volcaniclastic Combia

22 Formation. At this stage it is not clear how the formation of this unit is related to arc

23 volcanism and which role the Nazca plate subduction beneath the western margin of

24 South America plays. The timing, petrography and geochemical characteristics of

25 Combia Formation rocks were studied in the western and eastern parts of the

26 Amagá basin, in order to gain more information on the type of magma generation

27 and volcanic activity that led to the deposition of the Combia Formation.

28 Apatite and zircon fission-track dating largely confirm a 12-6 Ma age for the

29 deposition of the Combia Formation. Petrographic and major element analyses show

30 that mainly trachy-andesite ignimbrites with a calc-alkaline composition were

31 deposited in the western Amagá basin, whereas the volcanic rocks of the eastern

1

32 Amagá basin are lava flow and fall-out deposits of basaltic andesites or of tholeiitic

33 composition. Trace element and isotopic analyses show that slab dehydration and

34 sediment melting were important in primary magma generation in the mantle wedge,

35 but the primary magma was mixed with lower continental crustal melts, resulting in

36 characteristic isotope signatures in the western and eastern Amagá basin. All this

37 points to subduction driven arc volcanism with slab dehydration, sediment melting

38 magma mixing.

39

40

41 Introduction

42 The late Paleogene to present-day magmatism of northwestern South

43 America can be divided into four major phases of activity at about 24-20 Ma, 12-6

44 Ma, 6-3 Ma, and 3 Ma to the Present (e.g. Sierra, 1994; Toro et al., 1999; Gonzalez,

45 2001; Ramírez et al., 2006; Cediel et al., 2011; Pérez et al., 2013; Lesage et al.,

46 2013). These different magmatic phases are related to the complex tectonic setting

47 in which the Caribbean, Nazca and South American plates interact with each other

48 (Fig. 1). The break-up of the Farallón plate into the Nazca and Cocos plates between

49 26 and 24 Ma (Marriner and Millward, 1984), the reorientation of subduction

50 direction (Pardo-Casas and Molnar, 1987), and collision of the Panamá-Choco block

51 with northwestern South America at about 25 Ma drove the first magmatic pulse (e.g.

52 McCourt et al., 1984; Aspden et al., 1987; Kellogg and Vega, 1995; Trenkamp et al.,

53 2002; Cediel et al., 2003; Lonsdale, 2005; Restrepo-Moreno et al., 2010; Farris et

54 al., 2011). Second, since the late Paleogene the Nazca plate subduction zone was

55 subjected to changes in subduction angle and direction over time, resulting in

56 Miocene-Pliocene magmatic intrusions in the Western and Central Cordillera and

2

57 deposition of the Combia Formation in the Amagá basin (e.g. Pardo-Casas and

58 Molnar, 1987; Taboada et al., 2000; Cediel et al., 2003; Vargas and Mann, 2013). At

59 the same time, subduction of the Caribbean plate beneath the northern (Caribbean)

60 margin of South America caused isolated late Miocene-Pliocene volcanic activity in

61 the Eastern Cordillera (e.g. Vargas and Mann, 2013), such as in the Vetas-California

62 gold-mining district of the Santander Massif (Mantilla et al., 2013), or the Paipa-Iza

63 complex 150 km to the north-east of Bogotá (Fig. 1; Padro et al., 2005; Bernet et al.,

64 2016). Today the main volcanic activity in Colombia is focused on the Central

65 Cordillera with for example the Nevado del Ruiz, Nevado del Tolima, Cerro Machín,

66 Nevado del Huila, Azufral, Cumbal, etc. well to the south of the study area (Fig. 1;

67 e.g. Marín-Cerón et al., 2010, 2019; Leal-Mejía, 2011).

68 Different techniques have been used for more than a century to understand

69 the genesis, age and evolution of the Combia Formation, including petrography,

70 heavy mineral analysis, X-ray diffraction, geochemistry, geochronology,

71 thermochronology and stratigraphic analyses (e.g. Grosse, 1926; Jaramillo, 1976;

72 Calle and González, 1980; Álvarez, 1983; Marriner and Millward, 1984; Rios and

73 Sierra, 2004; Pérez, 2005; López et al., 2006; Ramírez et al., 2006), but the

74 evolution of the Nazca plate subduction zone magmatism still remains poorly

75 constrained. Here we present a study of a suite of samples collected from three

76 sections, the Cerro Amarillo section in the eastern Amagá basin, and the Anzá-

77 Bolombolo and La Metida Creek sections in the western Amagá basin (Fig. 2), in

78 order to improve the knowledge gained so far about the Combia Formation. The

79 volcaniclastic, tuff/lapilli and flow deposits of the Combia Formation were examined

80 with a) apatite fission-track (AFT) and zircon fission-track (ZFT) thermochronology,

81 b) petrographic analyses, and c) major and trace element analysis, as well as Sr, Nd

3

82 and Pb isotope analyses. All this was done with the objective of a) characterizing

83 and comparing the eastern and western volcanic deposits, and b) to better

84 understand the mid-late Miocene evolution of the Nazca subduction zone

85 magmatism manifested between the Western and Central Cordilleras.

86

87 Geological setting

88 The Northern Andes of northwestern South America consist in Colombia of

89 the Western, Central and Eastern Cordilleras (Fig. 1). Each of these mountain belts

90 reflects a particular part of the long-term evolution of the Northern Andes, which is

91 characterized by magmatic episodes since the Precambrian, during the Triassic,

92 Jurassic, Late Cretaceous, and since the late Paleogene/Neogene until today

93 (Aspden et al., 1987). In general, these magmatic phases have been related to

94 Farallón/Nazca plate subduction beneath the western margin of the South American

95 plate (e.g. Marriner and Millward, 1984; McCourt et al., 1984; Cediel et al., 2003;

96 Saenz, 2003; Restrepo-Moreno et al., 2009; Rodríguez et al., 2012). Accretion of

97 tectonic blocks or terranes of oceanic affinity to the continental margin during the late

98 Mesozoic and early Cenozoic did not cause Andean-type subduction volcanism,

99 because of their relatively young age and high buoyancy preventing subduction

100 (Cediel et al., 2003), and forcing surface uplift and the formation of the Western and

101 Central Cordilleras during the Pre-Andean and Andean orogenies (e.g. Van der

102 Hammen, 1960; Taboada et al., 2000; Cediel et al., 2003).

103 The present-day Andean volcanism is commonly divided into four volcanic

104 zones, the Northern Volcanic Zone (NVZ), Central Volcanic Zone (CVZ), Southern

105 Volcanic Zone (SVZ), and Austral Volcanic Zone (AVZ) (e.g. Thorpe and Francis,

106 1979; Thorpe et al., 1982; Stern, 2004; Marín-Cerón et al., 2019). These segments

4

107 have been distinguished based on differences in petrographic features and

108 geochemical signatures, and they are separated from each other by volcanic gaps

109 (e.g. Thorpe and Francis, 1979; Stern, 2004). The NVZ is located in north-western

110 South America and encompasses the region of present-day volcanism in the

111 Northern Andes of Ecuador and Colombia.

112

113 Geology of the Amagá basin

114 The Amagá basin forms the northern part of the much larger Amagá-Cauca-

115 Patía basin located between the Western and Central Cordilleras of the Northern

116 Andes in western Colombia (Fig. 1; Sierra and Marín-Cerón, 2011). Dextral strike–

117 slip faulting along the Cauca and Romeral fault systems to the west and east

118 respectively is responsible for development of the Amagá basin, which is tectonically

119 a pull – apart basin (e.g. Cediel et al., 2003). Basin evolution started possibly during

120 the Eocene (?) – Oligocene, with surface uplift and erosion of the Central Cordillera

121 from the Late Cretaceous to Eocene and deposition of clastic sediments of the

122 Lower Amagá Formation in the basin (e.g. Restrepo-Moreno et al., 2009). The Lower

123 Amagá Formation is known for its quartz-rich sandstones and mainly sub-bituminous

124 but locally anthracite grade coal (Silva et al., 2008; Blandon et al., 2008). The Lower

125 Amagá Formation is separated from the Oligocene to Miocene Upper Amagá

126 Formation by an unconformity and a change to a lithic arenite composition with

127 sedimentary and metamorphic lithoclasts derived from the Central Cordillera (Paez

128 Acuna, 2012). During the mid to late Miocene, subduction of the Nazca plate below

129 the South American plate allowed the genesis of the Combia Formation in the

130 Amagá basin (e.g. Grosse, 1926; Marriner and Millward, 1984; González, 2001;

131 Cediel et al., 2003; Ramírez et al., 2006; Leal-Mejía, 2011; Cediel et al., 2011).

5

132 Therefore, the Upper Amagá Formation is overlain by volcanic and volcaniclastic

133 deposits of the Combia Formation. Here we focus on the Cerro Amarillo section in

134 the Eastern Amagá basin and the Anzá – Bolombolo and Le Metida Creek sections

135 in the western Amagá basin.

136

137 The Cerro Amarillo section

138 This section is located between the towns of Damasco and La Pintada (Fig.

139 2). It has a total thickness of 193 m and comprises 34 layers of welded tuff,

140 pyroclastic and agglomerate breccia, lapilli tuff breccia, basalt and scoria (Fig. 3).

141 The layers vary in thickness from a 20 m pyroclastic and agglomerate breccia to 0.3

142 m lapilli tuff breccia, both at the top (Mesa-Garcia, 2015). There is also a 19.2 m

143 thick welded tuff at the bottom. However, it is most common to find layers of 1 – 7 m

144 in thickness. The layers are characterized by a tabular geometry. No evidence of

145 pinchout or lenses were observed in the outcrops. The bottom of the stratigraphic

146 succession is mainly characterized by lava flows and welded tuff while, the top of the

147 sequence mainly consists in coarse to very coarse grained pyroclastic flows. The

148 bottom layers have a strike and dip of S05°E/25°SW in average; towards the middle

149 of the section the layers strike and dip N70°E/18°SE. Finally the top breccia and

150 lapilli tuff layers strike and dip N15°W/19SW. Many basalt and welded tuffs layers

151 have randomly distributed vesicles and amygdules of variable size shapes, some are

152 elongated indicating lava flow directions.

153

154 The Anzá – Bolombolo section

155 This section is also located on the western bank of the Cauca River (Fig. 2)

156 between the villages of Anzá and Bolombolo. The stratigraphy of the 11.04 m thick

6

157 succession consists of tuffs and lapilli tuffs in the lower part of the section, which are

158 separated laterally from andesitic basalts and ash flow deposits by an erosional

159 unconformity (Fig. 4; Grosse, 1926; González, 2001; Sierra and Marín-Cerón, 2011).

160 In the upper part of the succession are a lapilli tuff breccia and a non-differentiated

161 lava flow, which cover the underlying units and the unconformity. No particular

162 sedimentary structures were observed at this location.

163

164 The La Metida Creek section

165 This section is located on the western bank of the Cauca River to the west of

166 Bolombolo (Fig. 2). The Combia Formation crops out along the stream bed (e.g.

167 González, 2001) and the exposed stratigraphic sequence has a thickness of 45 m

168 and is composed of 33 layers of tuff, lapilli tuff, lapilli tuff breccia, pyroclastic flows

169 and volcano-clastic sandstones (Fig. 5). Sedimentary structures found throughout

170 the sequence include lenticular bedding, load structures, ripple lamination and cross-

171 bedding. The layers at the bottom are mainly grain-supported, whereas the top

172 layers are matrix-supported. In addition, nodules and organic matter are commonly

173 observed at the bottom of the section, and the middle section is characterized by

174 bioturbation and fossilized plants (e.g. leaves). No such material was observed

175 towards the top of the section.

176

177 Analytical methods

178 Apatite and zircon fission-track analyses

179 AFT analysis was done on two samples from the AB section and six samples

180 from the MC section. In addition, ten samples of the MC section were analyzed with

7

181 the ZFT method. Unfortunately, the apatite and zircon yield of the CA section

182 samples was too low for fission-track analyses.

183 Sample preparation and analyses were performed at the thermochronology

184 laboratory of the Universidad EAFIT at Medellín and the ISTerre thermochronology

185 laboratory at the Université Grenoble Alpes. Rock samples were crushed and sieved

186 and heavy mineral fractions were separated using standard hydraulic, magnetic and

187 heavy liquid separation techniques. The apatite crystals were mounted in epoxy and

188 the zircon crystals in Teflon® sheets, polished and etched. Apatite grains were

189 etched for 20 seconds at 21°C with 5.5 mol HNO3, and zircons were etched at 228°C

190 for 10-40 h in a NaOH-KOH melt to reveal fission tracks. A white mica sheet was

191 mounted as the external detector. All samples were irradiated with thermal neutrons

192 at the FRM II reactor in Garching, Germany, with a nominal fluence of 8x1015 n/cm2

193 for apatite and 0.5x1015 n/cm2 for zircon, together with IRMM540R dosimeter glasses

194 and Durango age standards for apatite, and IRMM541 dosimeter glasses and Buluk

195 and Fish Canyon Tuff age standards for zircon. After irradiation external detectors

196 were etched in 48% HF for 18 min at 21°C. Fission tracks were counted dry at 1250x

197 using an Olympus BX51 microscope and the FTStage 4.04 system. Fission-track

198 ages for each sample were calculated using the Binomfit software of Brandon (see

199 Ehlers et al., 2005) and the RadialPlotter program of Vermeesch (2009).

200

201 Petrographic analyses

202 Petrographic analyses were performed on twenty samples, seven from the

203 Cerro Amarillo section, three from the Anzá-Bolombolo section, and ten from the La

204 Metida Creek section. Petrographic thin sections were prepared at Geoensayos

8

205 S.A.S., Medellín, Colombia. Some of the pyroclastic samples had to be impregnated

206 with epoxy, as these deposits were not well consolidated.

207 The samples were analyzed using an Olympus BX41TF petrographic

208 microscope at the Geology department of EAFIT University. Modal analysis was

209 performed counting 300 – 500 points per sample, using a point counter. The

210 description of mineral assemblages and textures was done according to Mackenzie

211 et al. (1984) and Ehlers (1987). The samples were classified using modal mineral

212 composition and the QAP diagram of Streckeisen (1976) for volcanic rocks and

213 based on the size of the material after Pettijohn (1975) for pyroclastic rocks.

214

215 Geochemical analyses

216 The rock samples collected in the field were crushed to 1 mm chips at the

217 Laboratory of Solid Materials at EAFIT University. Two hundred grams per sample

218 were separated thoroughly, choosing the rock chips that were the least weathered

219 and geochemical analyses for major and trace elements were performed in the clean

220 laboratory at the Institut des Sciences de la Terre (ISTerre) – Université Grenoble

221 Alpes, France. Samples were finely powdered in an agate mortar previous to

222 analyses, except for Pb isotope analysis for which rock chips were used directly. The

223 sample preparation and analytical procedures were executed according to Chauvel

224 et al. (2011), as summarized below.

225 For major elements, 50 mg of powdered sample were dissolved in 800 µl of

226 concentrated HNO3 and 15 drops of concentrated HF and heated in a Savillex

227 beaker for two days at 90°C on a hot plate. After cooling, 20 ml of H3BO3 (25 g/l)

228 were added to the solution to neutralize excess HF, 10 g of HNO3 and 250 ml of

229 milliQ water for further dilution. Five standards (BR 24, BEN, BHVO2, AGV–1 and

9

230 BCR–1), a duplicate and a blank were as well prepared for analysis. The solutions

231 were analyzed using Inductively Coupled Plasma Atomic Emission Spectroscopy

232 (ICP AES) at ISTerre to determine the major element composition of each sample.

233 Concentrations were obtained using the international rock standard BR to calibrate

234 the signal and the values recommended by Chauvel et al. (2011). Loss on ignition

235 (LOI) was calculated for all samples by heating one gram of sample at 1000°C for

236 one hour.

237 Sixteen samples, five standards (BR 24, BEN, BHVO2, AGV–1 and BCR–1),

238 three duplicates and one blank were analyzed for trace element contents. One

239 hundred milligrams of samples and standards were dissolved in 20 drops of

240 concentrated HNO3 and 3 ml of HF for three days at 120°C in Savillex beakers.

241 Samples were further taken up in 7N HNO3 and finally dissolved in 2% HNO3 with

242 traces of HF, to obtain a dilution factor of about 5000 (Chauvel et al., 2011). The

243 sample was analyzed using an Inductively Coupled Plasma Mass Spectrometry (ICP

244 MS) Agilent 7500 at ISTerre.

245 Ten samples were prepared for Nd, Pb and Sr isotope ratios analyses. Two

246 sets of Sr samples, unleached and leached, were prepared following the same

247 procedures as for the Nd samples and Pb samples, respectively. The samples were

248 dissolved and conditioned in Savillex beakers. Rock chips were leached according to

249 McDonough and Chauvel (1991), to eliminate as much as possible Sr and Pb

250 superficial contamination. For Nd and unleached Sr isotopic measurements, 100 mg

251 of sample were dissolved using HNO3 and HF and isolated using the same

252 procedure as Chauvel et al. (2011). For Pb and leached Sr, 1 g of rock chips was

253 leached prior to dissolution using HCl and then isolated as in Chauvel et al. (2011).

254 Nd and Pb isotopic ratios were measured using a Nu Plasma HR MC–ICPMS at

10

255 ENS Lyon, France, while Sr isotopic ratios were measured using a Thermo Scientist

256 Triton MS at the University of Brest, France.

257

258 Results

259 Fission-track results

260 Anzá – Bolombolo section

261 The tuff layer at the bottom of the Anzá-Bolombolo section (sample JJ22) has

262 an AFT central age of 8.4±3.1 Ma, and the pyroclastic flow sampled towards the top

263 of the section (sample JJ13) has an AFT central age of 7.9±1 Ma (Table 1; see the

264 supplementary data for individual grain ages and radial plots of all samples). The

265 combined Combia Formation AFT data are shown in a radial plot in Figure 6A with a

266 central age of 8.6±1.4 Ma based on 302 grain ages.

267

268 La Metida Creek section

269 In total six samples were analyzed with the AFT method and ten with the ZFT

270 method. AFT central ages range between 15.9±11.1 and 5.1±2.5 Ma. Nonetheless,

271 inherited single grains with apparent cooling ages of between 374 and 49 Ma can

272 also be observed in all samples (Table 1). Zircon crystals were mainly found in the

273 middle and bottom of the La Metida Creek section. The euhedral to subhedral zircon

274 crystals range in color from colorless, yellow, pink to red. The ZFT central ages are

275 between 12.7±2.4 Ma and 6.1±1.1 Ma (Table 2). Except for the volcaniclastic

276 sandstone sample (JJ6), no strong evidence exist of contamination with inherited

277 zircons derived from surrounding country rock (see the supplementary data for

278 individual grain ages and radial plots of all samples). The combined Combia

11

279 Formation ZFT data are shown in Figure 6B with a central age of 9.1±1.1 Ma based

280 on 346 grain ages.

281

282 Petrographic results

283 Cerro Amarillo section

284 In general, the seven samples analyzed for this section share similar

285 characteristics being hypocrystalline, porphyritic rocks. The mineral assemblage is

286 represented by plagioclase + pyroxene ± amphibole ± olivine (Table 3). Other

287 minerals present in the samples are secondary calcite, biotite and opaque minerals.

288 The rock fragments found in the ignimbrites are mainly basalts (aphanitic textures

289 and volcanic glass) or andesites (plagioclase crystals in volcanic glass). Some of

290 these fragments are fractured. There is also evidence of oxidation and sericitization,

291 even though carbonate minerals are the main alteration product in these samples.

292 The ignimbrites are dominated by crystals and rock fragments ranging from

293 55 – 70 %, whereas the two basalt samples have different percentage relationships

294 between the matrix and phenocrysts, CA – 14 rich in crystals and CA – 18 rich in

295 matrix. The crystal and/or rock fragments are in-equigranular, which is observed in

296 the presence of seriate, glomeroporphyritic, poikilitic, and interstitial textures.

297 Overgrowth textures such as skeletal, corona and crystal zoning are also found in

298 the samples, mainly in the ignimbrites. The crystals have subhedral to anhedral

299 shapes in most samples; euhedral-shaped crystals are rare or unable to identify due

300 to alteration. The matrix is generally altered to secondary calcite, but in some cases

301 a non-altered volcanic glass composition is observed with some embedded

302 microlites of plagioclase and less common pyroxene.

12

303 Based on the lack of quartz and potassic feldspar, all samples from the Cerro

304 Amarillo section plot in the basalt/andesite field of the classification diagram of

305 Streckeisen (1976) shown in Figure 7A. Color plates of thin section photographs are

306 given in the supplementary data archive.

307

308 Anzá – Bolombolo section

309 The analyzed samples are hypocrystalline, un-equigranular volcanic rocks.

310 The mineral assemblages are mainly plagioclase + pyroxene (Table 3). The matrix

311 consists of volcanic glass. Other rock components are spherulites, found in samples

312 AB–6 and AB–7, and olivine, found in sample AB–6. These samples are located to

313 the east of the unconformity described in the stratigraphic section. The samples are

314 characterized by having skeletal and crystal zoning overgrowth, spherulites and

315 vesicular textures. The lava flows can be classified as basalt/andesite in the

316 Streckeisen (1976) diagram (Fig. 7a), and the pyroclastic flow deposits as crystal

317 tuffs after Pettijohn (1975), as shown in Figure 7b. Color plates of thin section

318 photographs are given in the supplementary data.

319

320 La Metida Creek section

321 Nine samples from this section mainly plot in the crystal tuff field of the

322 Pettijohn (1975) classification (Fig. 7b). Only one sample (QML – 9) is defined as

323 tuffaceous sandstone because of the presence of cross-bedding in the layer.

324 Petrographic characteristics of all samples are similar and the classification as

325 pyroclastic and epiclastic rocks is based on sedimentary structures observed in the

326 field. The mineral assemblage is mainly plagioclase + pyroxene. Some amphiboles

327 are present at the bottom layers.

13

328 The sampled rocks are mainly hypocrystalline due to the presence of

329 phenocryst, and microlites in the matrix. The minerals have un-equigranular and

330 seriate fabrics. In general, skeletal textures are found in all analyzed samples. Other

331 common textures are glomero-porphyritic and crystal zoning, although the latter is

332 not present in the samples analyzed from the middle section of the La Metida Creek

333 section. Similar to the Anzá-Bolombolo section spherulites are found in the samples

334 from the bottom layers. Color plates of thin section photographs are given in the

335 supplementary data.

336

337 Geochemical results

338 Mayor element analyses of 16 samples are presented in Table 4. Loss on

339 ignition (LOI) is generally below 3%. For the Cerro Amarillo section samples, SiO2

340 contents range between 51 and 53 wt%; they plot in the tholeiitic field of the AFM

341 diagram (Fig. 8A) and correspond to basaltic andesite according to Figures 8B.

342 Samples coming from the Anza-Bolombolo area have slightly higher SiO2 content at

343 about 56 wt%; they plot in the alkaline field in Fig. 8A and correspond to trachy-

344 andesites in Fig. 8B.

345 Trace element contents of all samples are given in Table 5 and plotted in

346 primitive mantle-normalized spider diagrams in Fig. 9A. The Cerro Amarillo samples

347 display an enrichment in large-ion lithophile elements (Rb, Ba, Cs, Sr) and a strong

348 depletion in Nb and Ta. The Anzá-Bolombolo samples are even more enriched in

349 Rb, Ba, Cs and Sr, and with similar depletion in Nb and Ta (Fig. 9A). The main

350 difference between the Cerro Amarillo and Anzá-Bolombolo samples is the Li

351 enrichment of the Anzá-Bolombolo section samples. Samples from both sections are

352 enriched in Light Rare Earth Elements (LREE) relative to the Heavy Rare Earth

14

353 Elements (HREE) (Fig. 9B) with a stronger fractionation in the Anzá-Bolombolo

354 section samples. No significant Eu anomaly exists for all samples.

355

356 Nd, Sr, and Pb isotope compositions

357 Nd, Sr, and Pb isotope analyses of all samples are given in Table 6. Strontium

358 isotopic ratios measured on leached and unleached samples are similar within

359 errors. 87Sr/86Sr ratios range from 0.703862 to 0.703931 for the Cerro Amarillo

360 section samples, but are higher at about 0.70417 for the Anzá-Bolombolo section

361 samples (Figure 10A). 143Nd/144Nd ratios vary between 0.51292 and 0.51298 for the

362 Cerro Amarillo section samples, and are somewhat lower at 0.51290 for the AB

363 section samples (Figure 10B and C). Finally, the Cerro Amarillo section samples

364 define a small range in Pb isotopic ratios (208Pb/204Pb: 38.68 to 38.80, 207Pb/204Pb:

365 15.61 to 15.62 and 206Pb/204Pb: 18.91 to 19.07), and the Anzá Bolombolo section

366 samples fall in the middle of the range (Figure 11a and b)

367

368

369 Discussion

370

371 Constraints from low-temperature thermochronology

372 The AFT and ZFT data presented in this study can in principle be used for

373 constraining the age of deposition of the volcanic and volcaniclastic deposits of the

374 Combia Formation (Kowallis et al., 1986; Bernet et al., 2016). The AFT and ZFT data

375 central age values ranging from 15.9 – 5.1 Ma for AFT and 12.7 – 6.1 Ma for ZFT,

376 bracket the known 12-6 Ma age of Combia Formation volcanic activity (e.g. Leal-

377 Mejía, 2011). Our fission-track data highlight two important aspects that are

15

378 challenging in dating relatively young volcanic and volcaniclastic deposits with the

379 fission-track method. As to be expected the ZFT data correspond more closely to the

380 known 12-6 Ma age range determined from zircon U-Pb analyses (Leal-Mejía, 2011),

381 because of the higher U concentration and the better track counting statistics

382 resulting in higher precision results. The AFT data suffer under the very low (in

383 general <10 ppm) U concentration of the apatites in the Combia Formation. As can

384 been seen in the single grain data provided in the supplementary data archive, many

385 apatite grains are zero-track grains, resulting in a very high single grain age

386 uncertainty, as the induced track counts also tend to be because of the low U

387 concentrations.

388 In addition, dealing with volcanic and particularly volcaniclastic deposits the

389 risk of contamination with apatites and zircons recycled from the country rock is high,

390 and has been shown to be the case for certain volcanic deposits of the Paipa-Iza

391 volcanic complex (Bernet et al., 2016). Here we also think that zircons with >12 Ma

392 apparent cooling ages were most likely recycled from the Amagá Formation

393 (Piedrahita et al., 2017). Pre-Miocene cooling ages are common in apatites of the

394 Combia Formation deposits (Table 1), and are considered to be derived from

395 underlying basement rock and recycling of the Amagá Formation.

396 In summary, based on the AFT and ZFT data, volcanic activity occurred

397 between 12 Ma and 6 Ma. The main phase of activity for these deposits may have

398 been at around 9 Ma, as suggested by the central ages of the combined AFT and

399 ZFT data sets shown in Figure 6. This confirms a late Miocene depositional age,

400 which has previously been proposed based on stratigraphic position and whole rock

401 K-Ar dating of hypabyssal porphyries (e.g. Grosse, 1926; Restrepo et al., 1981;

16

402 Marriner and Millward, 1984; González, 2001, Pérez, 2005; Ramírez et al., 2006;

403 Leal-Mejía, 2011).

404

405 Shallow-level processes prior to eruption

406 The results presented in this study indicate that there are two different

407 petrographic trends among the Combia Formation rocks. In general, the rocks have

408 porphyritic textures, and plagioclase is one of the main mineral components, both as

409 phenocrysts and microlites. Similar results have been reported by Marriner and

410 Millward (1984), López et al. (2006) and Ramírez et al. (2006).

411 The matrix for most samples is comprised of volcanic glass with microlites of

412 plagioclase and to a lesser extent pyroxene. Devitrification of the volcanic glass

413 matrix is common in the samples. The samples show evidence of alteration (e.g.

414 secondary calcite, oxidation, argillitization) which may indicate metasomatic to

415 surficial processes related to hydrothermal alterations and mineralization processes

416 in the hypabyssal porphyries of the study area (e.g. Tassinari et al., 2008; Leal-

417 Mejía, 2011; Lesage et al., 2013; Uribe-Mogollón, 2013).

418 All samples show disequilibrium textures such as skeletal, sieve and

419 spherulitic textures, embayments, reaction rims, coronas and crystal zoning. These

420 textures are attributed to different conditions and magmatic processes such as

421 pressure variations, zoned magma chambers, decompression, magma mixing,

422 phenocrysts recycling and fractional crystallization (e.g. Nixon and Pearce, 1987;

423 Nelson and Montana, 1992; Singer et al., 1995; Perugini et al., 2003; Aldanmaz,

424 2006; Maro and Remesal, 2011). Thus, our petrographic results favor mainly magma

425 mixing and fractional crystallization to be the main causes for disequilibrium in the

426 magma chamber.

17

427

428 Magma genesis

429 Magmatic arcs are the result of subduction of oceanic crust beneath

430 continental crust and this complex setting has impact on magmatic processes such

431 as partial melting, fractional crystallization, changes in pressure and temperature,

432 sediment melting, dehydration, and decarbonation (e.g. Rollinson, 1993; Albarade,

433 1995; White, 2013). In addition, fluids added to magma play an important role, as

434 fluids allow the transport of incompatible elements from the subducted oceanic

435 basalt and sediments to the magma, leading to enrichment in specific mobile

436 elements and modifying the isotopic composition of the asthenospheric magma

437 source (Tatsumi, 2005; Tatsumi and Stern, 2006; Tatsumi and Takahashi, 2006;

438 Nakamura et al., 1985).

439 Our new geochemical data for the Cerro Amarillo area are generally similar to

440 what was previously published by Marriner and Millward (1984) and by Ordoñez

441 (2001). Major elements allow identification of a tholeiitic trend in the samples from

442 the eastern Amagá basin while the samples from the western Amagá basin follow a

443 calc-alkaline trend (Fig. 8A). Both magma suites have been previously recognized at

444 other sampling sites of the Combia Formation (Fig. 8A; e.g. Álvarez, 1983; Marriner

445 and Millward, 1984; Ordoñez, 2001; Leal-Mejía, 2011). With SiO2 contents between

446 47 – 59 wt% our Cerro Amarillo samples plot in the basaltic andesite field of Fig. 8B

447 and the samples from Anzá Bolombolo plot in the trachy-andesite field of LeMaitre et

448 al. (1989) and Cox et al. (1979) (Fig. 8B). For comparison, published data for the

449 Combia Formation are also shown in this figure. It appears clearly that the AB

450 section samples from the western Amagá basin are much more alkaline than others

451 (Fig. 8B), and that samples from both sections have lower SiO2 contents than the

18

452 volcanic rocks of the 24-20 Ma magmatic phase and most of the 17-9 Ma magmatic

453 phase rocks analyzed by Leal-Mejía (2011). Despite these differences, so far no

454 precise division of the basin with respect to the magmatic suites can be given at the

455 moment, and more detailed mapping and geochemical analyses are necessary.

456 The trace elements provide clear evidence of a subduction-related

457 geochemical signature, with Nb and Ta depletion (e.g. Wilson, 1989; White, 2013).

458 Samples also all have very high Ba, U, Pb and Sr contents as already noticed for

459 older samples from the Combia Formation (Fig. 9C, Leal-Mejía, 2011 data

460 summarized in Marín-Cerón et al., 2019). Medium to slightly elevated Ba/Th values

461 (a proxy for slab dehydration Fig. 9E; Labanieh et al., 2012) characterize the Cerro

462 Amarillo samples from the eastern Amagá basin, whereas the AB samples of the

463 western Amagá basin have low ratio consistent with a sediment melting trend.

464 In general, the trace element results of the Amagá basin resemble those of

465 the northern volcanic zone (NVZ) of the Andes, as summarized in Marín-Cerón et al.

466 (2019), even though the Combia Formation has higher LILE (Rb, Sr, Ba) contents

467 and lower Nb and Ta contents than those reported for the NVZ (Thorpe et al., 1982;

468 Marriner and Millward, 1984).

469 The REE patterns of samples from the Anza Bolombolo and Cerro Amarillo

470 sections are somewhat different, with higher fractionation in the former group (Fig.

471 9b). In comparison to the REE spectra of the 17-6 Ma porphyritic intrusions and

472 recent volcanism, the Combia Formation shows a pattern closer to the latest stage of

473 volcanic activity (Fig. 9D; data from Leal-Mejía (2011), summarized by Marín-Cerón

474 et al. (2019)). The difference in the slope of the REE patterns of eastern and western

475 Amagá basin samples suggest differences in terms of magma formation, most

476 probably during melting processes in the mantle source. Overall, the absence of Eu

19

477 anomaly suggests that melting occurred relatively deep, below the plagioclase

478 stability level (~40 km depth).

479

480 Magma source composition

481 Combining trace element data with isotopic data can help understanding the

482 origin of magmas and the potential role of subducted slab addition to the mantle

483 wedge (e.g Tatsumi, 2005; Tatsumi and Stern, 2006). The Cerro Amarillo and Anzá-

484 Bolombolo samples have medium to slightly elevated Sr/Th and rather low 87Sr/86Sr

485 ratios (Fig. 10A), indicating transfer of elements from the subducted slab byNVZ

486 aqueous fluids. Because no systematic difference in 87Sr/86Sr is observed between

487 leached and unleached samples (Table 6), post-emplacement alteration can be

488 excluded, and the measured ratios can be considered as representative of the

489 magma source (e.g. Tamura and Nakamura, 1996; Shibata and Nakamura, 1997).

490 Similarly, the large range of Ba/Th at almost constant 143Nd/144Nd indicates addition

491 of Ba through slab dehydration (Fig. 10B), with a stronger effect in the eastern

492 Amagá basin than in the western part of the basin. In a 87Sr/86Sr vs 143Nd/144Nd

493 isotopic space (Fig. 10C), the Amaga basin samples fall as expected within the field

494 defined by the NVZ (see compilation of Marín-Cerón et al., 2019). It is worth noting

495 that the Amaga Basin samples correspond to the most depleted part of the isotopic

496 range observed along the Andes, suggesting limited crustal contamination during

497 magma ascent and/or limited contribution of subducted sedimentary material to the

498 mantle wedge. It is also worth noting that the Anzá-Bolombolo samples have slightly

499 more enriched isotopic characteristics than the Cerro Amarillo samples, a feature

500 consistent with the difference seen in Figure 9E since sediment addition tends to

501 lower the Nd isotopic composition and increase the Sr isotopes.

20

502 Lead isotopes provide complementary and useful information. As was the

503 case for Sr and Nd isotopic data, the Amaga basin samples fall in the field defined by

504 the northern volcanic zone and differ drastically from fields defined by central and

505 southern volcanic zones (Fig. 11A) (see compilations made by Marín-Cerón, 2007

506 and Marín-Cerón et al., 2010; 2019). The rather radiogenic values of Pb isotopes for

507 the northern volcanic zone have been interpreted as being due either to an enriched

508 mantle reservoir or to continental crust assimilation. The enriched mantle hypothesis

509 was suggested by Rodríguez-Vargas et al. (2005) on the basis of the Nd and Sr

510 isotopic characteristics of xenoliths from the Mercaderes region in SW Colombia.

511 The authors invoked the potential involvement of magma material coming from the

512 Galapagos plume, but such influence under the Amaga basin seems quite unlikely

513 given the distance between the SW Colombian arc and the Galapagos (>450 km,

514 Pedersen and Furnes, 2001). However, the Amaga Basin samples analyzed in this

515 study provide new information because they define a tight correlation in 208Pb/204Pb

516 vs 206Pb/204Pb space (see Fig 11B). Such linear array implies the involvement and

517 mixture of two endmembers whose compositions remain unchanged during the

518 entire volcanic sequence. The enriched end-member seems to correspond to the

519 local lower continental crust (see Fig. 11B) while the less radiogenic endmember is

520 more ambiguous. Following Marin-Ceron’s model (2019), this ‘depleted’ endmember

521 could correspond to the mantle wedge whose composition would be affected by the

522 presence of material originating from the subducted slab (see Fig. 11B).

523 Nonetheless, there might be differences in the subduction mechanisms, proportion of

524 end-members interaction, and magma source evolution between magmatism at 12 –

525 6 Ma and magmatism at 3 Ma – present, despite assuming the same end-members.

526 In summary, magma was generated by slab dehydration, sediment melting and

21

527 interactions with the LCC, resulting in the mixing of at least two end-member

528 sources. Due to differences in depth of melting and other magmatic processes (AFM

529 and MASH), as shown in Figure 12, the late Miocene volcanic rocks of the western

530 and eastern Amagá basin show distinct petrologic and geochemical signatures.

531

532 Conclusions

533 The late Miocene volcanic and volcanoclastic Combia Formation of the

534 Amagá basin between in the Central and Western Cordillera developed above the

535 Nazca plate subduction zone in western Colombia. Volcanic and volcaniclastic rocks

536 of the Combia Formation are characterized by a sequence of ignimbrites and lava

537 flows of tholeiitic affinity at the bottom and pyroclastic flows to the top of the

538 formation in the eastern Amaga basin and a succession of pyroclastic and epiclastic

539 flows with calc-alkaline affinity in the western Amagá basin. Using apatite and zircon

540 fission-track dating the timing of volcanic activity during the deposition of the Combia

541 Formation was confirmed between 12 and 6 Ma.

542 Trace element, REE and Nd, Sr and Pb isotopic analyses show that the

543 surface weathering did not modify the geochemical signatures and that the

544 geochemical composition of the samples results of several magmatic processes

545 including slab dehydration and sediment melting to form the primary magma in the

546 mantle wedge and mixing of this primary magma with lower continental crustal

547 before eruption. In contrast, contamination by upper crustal rocks could not be

548 detected. Finally, the new geochemical results confirm that volcanism in the Combia

549 area between 12 and 6 Ma was similar to what is known about the NVZ of the Andes

550 in South America.

551

22

552

553 Acknowledgements

554 We acknowledge support of this study from ECOS-NORD/Colciencias/ICETEX 555 project C12U01 of M. Bernet and M.I. Marín-Cerón, as well as a BQR SUD grant at 556 ISTerre, awarded to M. Bernet. We thank Wilton, Francois Senebier and Francis 557 Coeur for help with sample preparation and mineral separation.

558

559

23

560 References

561 Albarade, F. 1995. Introduction to geochemical modelling. Cambridge University

562 Press. 543 pp.

563 Aldanmaz, E. 2006. Mineral – chemical constraints in the Miocene calc-alkaline and

564 shoshonitic volcanic rocks of Western Turkey: Disequilibrium phenochryst

565 assemblages as indicators of magma storage and mixing conditions. Turkish Journal

566 of Earth Sciences. 15: 47 – 73.

567 Álvarez, A. 1983. Geología de la cordillera Central y el Occidente colombiano y

568 petroquímica de los intrusivos granitoides Mesocenozoicos. Boletín Geológico. 26:

569 175 pp.

570 Aspden, J.A., McCourt, W.J., and Brook, M., 1987. Geometrical control of

571 subduction-related magmatism: the Mesozoic and Cenozoic plutonic history of

572 Western Colombia. Journal of the Geological Society, London. 144: 893 – 905.

573 Barret, T.J. 1983. Strontium- and lead-isotope composition of some basalts form

574 Deep Sea Drilling Project Hole 504B, Costa Rica Rift, Legs 69 and 70. In Cann, J.R.,

575 Langseth, M.G., Honnorez, J., VonHerzen, R.P., White, S.M., et al., Init. Repts.

576 DSDP, 69. Washington (U.S. Govt. Printing Office). 643 – 650.

577 Barret T. J., Taylor P. N., Lugowski J. 1987. Metalliferous sediments from DSDP leg

578 92: the East Pacific Rise transects. Geochim. Cosmochim. Acta. 46:651 – 666.

579 Bernet, M., Uruena, C., Amaya, S.,Pena, M.L., 2016. New thermo- and

580 geochronological constraints on the Pliocene-Pleistocene eruption history of the

581 Paipa-Iza volcanic complex, Eastern Cordillera, Colombia, Journal of Volcanology

582 and Geothermal Research , 327:299-309.

24

583 Blandon, A., Parra, N., Gorin, G.E., Arango, F. 2008. Adapting palynological

584 preparation methods in subbituminous and bituminous coals from Colombia to

585 improve palynofacies and hydrocarbon source rock evaluations. International Journal

586 of Coal Geology, 73:99–114.

587 Calle, B., González, H. 1980. Geología y Geoquímica de la Plancha 166, Jericó.

588 Informe no. 1822. Medellín. INGEOMINAS. 232 pp.

589 Cediel, F., Shaw, R.P., Caceres, C. 2003. Tectonic assembly of the Northern

590 Andean Block, in C. Bartolini, R. T. Buffler, and J. Blickwede, eds., The Circum-Gulf

591 of Mexico and the Caribbean: Hydrocarbon habitats, basin formation, and plate

592 tectonics.- AAPG Memoir 79: 815-848.

593 Cediel, F., Leal-Mejía, H., Shaw, R.P., Melgarego, J.C., Restrepo-Pace, P.A. 2011.

594 Petroleum Geology of Colombia: Regional Geology of Colombia. ANH – Colombia.

595 1: 220 pp.

596 Chauvel, C., Bureau, S., Poggi, C. 2011. Comprehensive Chemical and Isotopic

597 Analyses of Basalt and Sediment Reference Materials. Geostandards and

598 geoanalytical research. 35 (1): 125 – 143.

599 Chiaradia, M., Fontbote, L. 2002. Lead isotope systematics of Late Cretaceous

600 Tertiary Andean arc magmas and associated ores between 8°N and 40°S: evidence

601 for latitudinal mantle heterogeneity beneath the Andes. Terra Nova. 14 (5): 337 –

602 342.

603 Cox, K.G., Bell, J.D., Pankhurst, R.J. 1979. The Interpretation of Igneous Rocks:

604 London, George Allen & Unwin. 464 pp.

25

605 Dasch, E.J. 1981. Lead isotopic composition of metalliferous sediments from the

606 Nazca Plate. Mem. Geol. Soc. Am. 154: 199 – 209.

607 Davidson, J.P., de Silva, S.L. 1992. Volcanic rocks from the Bolivian Altiplano:

608 Insights into crustal structure, contamination, and magma genesis in the central

609 Andes. Geology. 20: 1127 – 1130.

610 de Silva, S. 1991. Styles of zoning in Central Andean ignimbrites-insights into

611 magma chamber processes: In: Andean magmatism and its tectonic setting. Special

612 paper 265: 217 – 232.

613 DePaolo, D. J. (1981). Trace-element and isotopic effects of combined wallrock

614 assimilation and fractional crystallisation. Earth and Planetary Science Letters 53,

615 189–202.

616 Ehlers, E.G. 1987. Optical Mineralogy: Vol. 2. Mineral Descriptions. Blackwell

617 Scientific Pub. Palo Alto. 286 pp.

618 Ehlers, T.A., Chaudhri, T., Kumar, S., Fuller, C., Willett, S.D., Ketcham, R., Brandon,

619 M.T. 2005.Computational tools for low-temperature thermochronometer

620 interpretation. In: Reiners, P.W., Ehlers, T.A. (Eds.), Low-temperature

621 Thermochronology. Techniques, Interpretations and Applications. Reviews in

622 Mineralogy and Geochemistry, Mineralogical Society of America 58:pp. 589–622.

623 http://dx.doi.org/10.2138/rmg.2005.58.22.

624 Evensen, N. M., Hamilton, P. J., O‟Nions, R. K. 1978. Rare-earth abundances in

625 chondritic meteorites. Geochimica et Cosmochimica Acta. 42: 1199 – 1212.

626 Farris, D.W., Jaramillo, C., Bayona, G., Restrepo-Moreno, S.A., Montes, C.,

627 Cardona, A., Mora, A., Speakman, R.J., Glascock, M.D., Valencia, V. 2011.

26

628 Fracturing of the Panamanian Isthmus during initial collision with South America.

629 Geology. 39 (11): 1007 – 1010.

630 Frey, F.A., Gerlach, D.C., Hickey, R.L., López-Escobar, L., Minizaga-Villavicencio, F.

631 1984. Petrogenesis of the Laguna del Maule volcanic complex, Chile (36°S). Contrib.

632 Mineral Petrol. 88: 133 – 149.

633 González, H. 2001. Memoria Explicativa del Mapa Geológico del Departamento de

634 Antioquia. Escala 1:400.000. Medellín, INGEOMINAS. 240 pp.

635 Grosse, E., 1926. Estudio Geológico del Terciario carbonífero de Antioquia en la

636 parte occidental de la Cordillera Central de Colombia: Berlín, Verlag Von Dietrich

637 Reimer. 361 pp.

638 Halliday, A.N., Lee, D-C., Tommasini, S., Davies, G.R., Paslick, C.R., Fitton, J.G.,

639 James, D.E. 1995. Incompatible trace elements in OIB and MORB and source

640 enrichment in the sub-oceanic mantle. Earth Planet. Sci. Lett. 133: 379 – 395.

641 Harmon, R.S., Barreiro, B., Moorbath, S., Hoefs, J., Francis, P.W., Thorpe, R.S.,

642 Deruelle, B., McHugh, J. and Viglino, J.A. 1984. Regional O-, Sr-, and Pb- isotope

643 relationships in late Cenozoic calc-alkaline lavas of the Andean Cordillera: Journal of

644 the Geological Society. 141 (5): 803 – 822.

645 Hawkesworth, C.J., Powell, M. 1980. Magma genesis in the Lesser Antilles island

646 arc. Earth and Planetary Science Letters. 51: 297 – 308.

647 Hawkesworth, C.J., Norry, M.J., Roddick, J.C., Baker, P.E., Francis, P.W., Thorpe,

648 R.S., 1979. 143Nd/144Nd, 87Sr/86Sr, and incompatible trace element variations in

649 calc-alkaline andesitic and plateau lavas from South America. Earth Planet. Sci.

650 Letters. 42: 45 – 57.

27

651 Hemming S.R, McLeannan, S.M. 2001. Pb isotopic compositions of modern deep

652 sea turbidites. Earth Planet. Sci. Let. 184, 489 – 503.

653 Hickey, R.L., Frey, F.A., Gerlach, D.C. López-Escobar, L. 1986. Multiple sources for

654 basaltic arc rocks from the southern volcanic zone of the Andes (34° - 41° S): Trace

655 element and isotopic evidence for contributions from subducted oceanic crust,

656 mantle, and continental crust. J. Geophys. Res. 91 (B6): 5963 – 5983.

657 Hildreth, W.; Moorbath, S. 1988. Crustal contribution to arc magmatism in the Andes

658 of central Chile. Contribution to Mineralogy and Petrology. 98: 455 – 489.

659 Irvine, T.N., Baragar, W.R.A. 1971. A guide to the chemical classification of the

660 common volcanic rocks: Canadian Journal of Earth Sciences. 8 (5): 523 – 548.

661 Ishikawa T., Nakamura E. 1994. Origin of the slab component in arc lavas from

662 across-arc variation of B and Pb isotopes. Nature. 370: 205 – 208.

663 James, D.E. 1982. A combined O, Sr, Nd, and Pb isotopic and trace element study

664 of crustal contamination in central Andean lavas: I. Local geochemical variations.

665 Earth and Planetary Science Letter. 57: 47 – 62.

666 James, D.E., Brooks, C., Cuyubamba, A. 1976. Andean Cenozoic volcanism:

667 Magma genesis in the light of strontium isotopic composition and trace-element

668 geochemistry. Geol. Soc. Amer. Bull. 87: 592 – 600.

669 Jaramillo, J. M., 1976. Volcanic rocks of the Río Cauca valley, Colombia S.A. Thesis

670 Degree of Master of Arts, Rice University, Houston.

671 Kay, S., Mpodozis, C., Ramos, V.A., Munizaga, F. 1991. Magma source variations

672 for mid-Tertiary magmatic rocks associated with a shallowing subduction zone and a

28

673 thickening crust in the Central Andes (28 – 33°S). In: Andean Magmatism and its

674 Tectonic Setting, Boulder, Colorado. Harmon, R.S., Rapela, C.W., eds. Spec. Pap.

675 Geol. Soc. Am. 265: 113 – 137.

676 Kellogg, J., Vega, V. 1995. Tectonic development of Panamá, Costa Rica, and the

677 Colombian Andes: Constraints from global positioning system geodetic studies and

678 gravity. Geological Society of America. Special Paper 295: 75 – 90.

679 Kerr, A.C. 2003. Oceanic Plateaus. In: Rudnick, R. (ed.) The Crust. Elsevier

680 Science: Oxford. Treatise on Geochemistry. 3: 537 – 565.

681 Kowallis, B.J., Heaton, J.S., Bringhurst, K., 1986. Fission-track dating of volcanically

682 derived sedimentary rocks. Geology 14, 19–22.

683 Labanieh, S., Chauvel, C., Germa, A., Quidelleur, X. 2012. Martinique: a clear case

684 for sediment melting and slab dehydration as a function of distance to the trench.

685 Journal of Petrology. 53: 2441 – 2464.

686 Leal-Mejía, H. 2011. Phanerozoic Gold Metallogeny in the Colombian Andes – A

687 tectono-magmatic approach: Ph.D. thesis, Barcelona (Catalonia), Spain, University

688 of Barcelona, 1000p.

689 LeMaitre, R. W., Bateman, P., Dudek, A., Keller, J., Lameyre-LeBas, M. J., Sabine,

690 P. A., Schmid, R., Sorensen, H., Streckeisen, A., Woolley, A. R. & Zanettin, B., A.

691 1989. Classification of Igneous Rocks and Glossary of Terms Oxford: Blackwell.

692 Lesage, G., Richards, J.P., Muehlenbachs, K., Spell, T.L. 2013. Geochronology,

693 geochemistry, and Fluid Characterization of the Late Miocene Buriticá Gold Deposit,

694 Antioquia Department, Colombia. Economic Geology. 108: 1067 – 1097.

29

695 Lonsdale, P. 2005, Creation of the Cocos and Nazca plates by fission of the

696 Farallón plate: Tectonophysics. 404 (3 – 4): 237 – 264

697 López, A., Sierra, G.M., Ramírez, S. 2006. Vulcanismo Neógeno en el suroccidente

698 antioqueño y sus implicaciones tectónicas. Boletín Ciencias de la Tierra. 19: 27 – 41.

699 Mackenzie, W.S., Donaldson, C.H., Guilford, C. 1984. Atlas of igneous rocks and

700 their textures. Longman Scientific and Technical, UK.

701 Mantilla, L.C., Bissig, T., Valencia, V., Hart, C.J.R., 2013. The magmatic history of

702 the Vetas–California mining district, Santander Massif, Eastern Cordillera, Colombia.

703 Journal of South American Earth Sciences 45, 235–249.

704 Marín-Cerón, M.I. 2007. Major, trace element and multi-isotopic systematics of SW

705 Colombian volcanic arc, northern Andes: Contributions of slab fluid, mantle wedge

706 and lower crust to the origin of Quaternary andesites. Doctoral thesis. Okayama

707 University, Japan. 133 pp.

708 Marín-Cerón, M.I., Moriguti, T., Makishima, A., Nakamura, E. 2010. Slab

709 decarbonation and CO2 recycling in the Southwestern Colombian volcanic arc.

710 Geochimica et Cosmochimica Acta. 74: 1104 – 1121.

711 Marín-Cerón, M.I., Leal-Mejía, H., Bernet, M., Mesa-Garcia, J., 2019. Late Cenozoic

712 to Modern-Day Volcanism in the Northern Andes: A Geochronological,

713 Petrographical, and Geochemical Review. In Cediel, F., Shaw, R.P. (eds.) Geology

714 and Tectonics of Northwestern South America, Frontiers in Earth Sciences, Springer

715 Nature Switzerland., p. 603-648.

30

716 Maro, G., Remesal, M.B. 2011. El vulcanismo de la Alta Sierra de Somún Curá: El

717 Cerro Corona, Provincia de Río Negro, Argentina. Abstract. XVIII Congreso

718 Geológico Argentino. Neuquén. 2 pp.

719 Marriner, G.F., Millward, D. 1984. Petrochemistry of Cretaceous to recent Vulcanism

720 in Colombia. Journal of Geological Society of London.141: 473 – 486.

721 McCourt, W.J., Aspden, J.A., Brook, M. 1984. New geological and geochronological

722 data from the Colombian Andes: continental growth by multiple accretion. Journal of

723 Geological Society, London. 141: 831 – 845.

724 McDonough, W.F., Chauvel, C. 1991. Sample contamination explains the Pb isotopic

725 composition of some Rurutu island and Sasha seamount basalts. Earth Planet. Sci.

726 Lett. 105: 397 – 404.

727 McDonough, W. F., Sun, S. S. 1995. Composition of the Earth. Chemical Geology.

728 120: 223 – 253.

729 Mesa-Garcia, J. 2015. Combia Formation: a Miocene immature volcanic arc? EAFIT

730 University Master thesis. 244 pp.

731 Nakamura, E., Campbell, I.H., Sun, S.S. 1985. The influence of subduction

732 processes on the geochemistry of Japanese alkaline basalt. Nature. 316: 55 – 58.

733 Nelson, S.T., Montana, A. 1992. Sieve-textured plagioclase in volcanic rocks

734 produced by rapid decompression. American Mineralogist. 77: 1242 – 1249.

735 Nixon, G.T., Pearce, T. H. 1987. Laser-interferometry study of oscillatory zoning in

736 plagioclase: The record of magma mixing and phenocryst recycling in calc- alkaline

737 magma chambers, Iztaccihuatl Volcano, Mexico. Amer. Mineral. 72: 1144 – 1162.

31

738 Ordoñez, O. 2002. Caracterizacao isotopica Rb-Sr E Sm-Nd dos principais

739 eventos magmáticos nos Andes Colombianos. Doctoral thesis. Brazilia University.

740 165 pp.

741 Páez Acuna, L. 2012. Análisis estratigráfico y de proveniencia del Miembro Superior

742 de la Formación Amagá en los sectores de la Pintada y Valparaiso (cuenca Amagá,

743 andes noroccidentales). Master thesis on Earth Sciences. EAFIT University.

744 Pardo, N., Cepeda, H., Jaramillo, J.M., 2005. The Paipa volcano, Eastern Cordillera

745 of Colombia, South America: volcanic stratigraphy. Earth Sci. Res. J. 9, 3–18.

746 Pardo-Casas, F., Molnar, P. 1987. Relative motion of the Nazca (Farallón) and

747 South American plates since Late Cretaceous time. Tectonics 6 (3): 223 – 248.

748 Patiño, L. C, Carr, M., Feigenson, M. 2000. Local and regional variations in Central

749 American arc lavas controlled by variations in subducted sediment input. Contrib.

750 Mineral Petrol. 138, 265-283.

751 Pedersen R., Furnes, H. 2001. Nd- and Pb-isotopic variations through the upper

752 oceanic crust in DSDP/ODP Hole 504B, Costa Rica Rift. Earth and Planetary Sc.

753 Lett. 189: 221 – 235.

754 Pérez, A. L., 2005. Formación Combia: Litofacies y aproximación a su edad con

755 base en estudios palinológicos, suroeste antioqueño. Undergraduate thesis.

756 Medellín, EAFIT University.

757 Pérez, A.M., Marín-Cerón, M.I., Bernet, M., Sierra, G., Moreno, N. 2013.

758 Resultados preliminares de AFT en la Formación Amagá, Pozos el Cinco- 1B y

759 Venecia-1. XIV Congreso Colombiano de Geología Libro: XIV Congreso

760 Colombiano de Geología. Resúmenes.

32

761 Perugini, D., Busa, T., Poli, G., Nazzareni, S. 2003. The role of chaotic dynamics

762 and flow fields in the development of disequilibrium textures in volcanic rocks.

763 Journal of Petrology. 44 (4): 733 – 756.

764 Pettijohn, F. J. 1975. Sedimentary rocks, third edition: New York, Harper & Row. 628

765 pp.

766 Piedrahita, V.A., Bernet, M., Chadima, M., Sierra, G.M., Marín-Cerón, M.I., Toro,

767 G.E., 2017. Detrital zircon fission-track thermochronology and magnetic fabric of the

768 Amagá Formation (Colombia): Intracontinental deformation and exhumation events

769 in the northwestern Andes. Sedimentray Geology, 356, 26-42.

770 Plank, T., Langmuir, C. 1998. The chemical composition of subducting sediment and

771 its consequences for the crust and mantle, Chem. Geol. 145, 325 – 394.

772 Ramírez, D.A., López, A., Sierra, G.M., Toro, G.E. 2006. Edad y provenincia de las

773 rocas volcánico sedimentarias de la Formación Combia en el suroccidente

774 Antioqueño- Colombia. Boletin Ciencias de la Tierra. 19: 9 – 26.

775 Restrepo, J.J., Toussaint, J.F., González, H., 1981. Edades MioPliocenas del

776 magmatismo asociado a la Formación Combia. Departamentos de

777 Antioquia y Caldas, Colombia. Geología Norandina. 3: 2126.

778 Restrepo-Moreno, S.A., Foster, D.A., Stockli, D.F., Parra-Sánchez, L.N. 2009. Long-

779 term erosion and exhumation of the „Altiplano Antioqueño‟, Northern Andes

780 (Colombia) from apatite (U-Th)/He thermochronology. Earth and Planetary Science

781 Letters. 278: 1 – 12.

782 Restrepo-Moreno, S.A., Cardona, A., Jaramillo, C., Bayona, G., Montes, C., Farris,

783 D.W. 2010. Constraining Cenozoic uplift/exhumation of the Panamá-Chocó Block by

33

784 apatite and zircon low-temperature thermochronology: insights on the onset of

785 collision and the morphotectonic history of the region. Abstract. GSA Denver Annual

786 Meeting. Geological Society of America Abstracts with Programs. 42 (5): 521.

787 Rios, A.M., Sierra, M.I. 2004. La Formación Combia: Registro de la relación entre el

788 volcanismo Neógeno y la sedimentación fluvial, sección Guineales – Bolombolo,

789 suroeste antioqueño. Undergraduate thesis. EAFIT University. 122 pp.

790 Rodríguez-Vargas, A., Koester, E., Mallmann, G., Conceicao, R.V., Kawashita, K.,

791 Weber, M.B.I. 2005. Mantle diversity beneath the Colombian Andes, Northern

792 Volcanic Zone: Constraints from Sr and Nd isotope. Lithos 82: 471 – 484.

793 Rodríguez, G., Arango, M.I., Bermúdez, J.G. 2012. Batolito de Sabanalarga,

794 plutonismo de arco en la zona de sutura entre las cortezas oceánica y continental de

795 los Andes del Norte. Boletín Ciencias de la Tierra. (32): 81 – 98.

796 Rollinson, H.R. 1993. Using geochemical data: evaluation, presentation,

797 interpretation. Longman group UK. 352 pp.

798 Saenz, E.A. 2003. Fission track thermochronology and denudational response to

799 tectonics in the north of the Colombian Central Cordillera. Master thesis. Shimane

800 Un Salters, V. & Stracke, A. 2004. Composition of depleted mantle, Geochem,

801 Geophys, Geosyst 5 (5): 27 pp.

802 Saunders, A.D., Norry M.J., Tarney, J. 1988. Origin of MORB and chemically

803 depleted mantle reservoirs: trace element constraints. J. Petrology, Special

804 Lithosphere Issue. 415 – 445.iversity. 131 pp.

805 Shibata T., Nakamura, E. 1997. Across-arc variations of isotope and trace element

806 compositions from Quaternary basaltic volcanic rocks in northeastern Japan:

34

807 implications for interaction between subducted oceanic slab and mantle wedge. J.

808 Geophys. Res. 102: 8051 – 8064.

809 Sierra, G. 1994. Structural and sedimentary evolution of the Irra Basin, northern

810 Colombian Andes. Master thesis, Department of Geological Science, State

811 University of New York, Binghamton, NY. 102 pp.

812 Sierra, G.M., Marín Cerón, M.I., 2011. Petroleum Geology of Colombia. Amagá,

813 Cauca and Patía Basins. Agencia Nacional de Hidrocarburos. 104 pp.

814 Silva, J.C., Sierra, G.M., Correa, L.G. 2008. Tectonic and climate driven fluctuations

815 in the stratigraphic base level of a Cenozoic continental coal basin, northwestern

816 Andes. Journal of South American Earth Sciences. 26: 369 – 382.

817 Singer, B.S., Dungan, M.A., Layne, G.D. 1995. Textures and Sr, Ba, Mg, Fe, K, and

818 Ti compositional profiles in volcanic plagioclase: Clues to the dynamics of calc-

819 alkaline magma chambers. American Mineralogist. 80: 776 – 798.

820 Stern, C.R. 2004. Active Andean volcanism: its geologic and tectonic setting.

821 Andean Geology. 31 (2): 161 – 206.

822 Stern, C.R., Moreno, H., López-Escobar, L., Clavero, J.E., Lara, L.E., Naranjo, J.A.,

823 Parada, M.A., Skewes, M.A. 2007. Chilean volcanoes. In: The Geology of Chile.

824 Moreno, T., Gibbons, W. (eds) The Geological Society, UK. 147 – 178.

825 Streckeisen, A. 1976. To Each Plutonic Rock its Proper Name. Earth Science

826 Review. 12: 1-33.

827 Sun, S.-s. and McDonough, W.F.. 1989. Chemical and isotopic systematics of

828 oceanic basalts: implications for mantle composition and processes. In: 4.D.

35

829 Saunders and M.J. Norry (Editors). Magmatism in the Ocean Basins. Geological

830 Society. London. pp. 3 13-345.

831 Taboada, A., Rivera, Luís A., Fuenzalida, A., Cisternas, A., Philip, Hervé, Bijwaard,

832 H., Olaya, J., and Rivera, C., 2000. Geodynamics of the northern Andes:

833 Subductions and intracontinental deformation (Colombia). Tectonics. 19 (5): 787 –

834 813.

835 Tamura, Y., Nakamura, E. 1996. The arc lavas of the Shirahama Group Japan: Sr

836 and Nd isotopic data indicate mantle-derived bimodal magmatism. Journal of

837 petrology. 37 (6): 1307 – 1319.

838 Tassinari, C.C.G., Díaz, F., Buena, J. 2008. Age and sources of gold mineralization

839 in the Marmato mining district, NW Colombia: A Miocene – Pliocene epizonal

840 gold deposit. Ore Geology Reviews. 33: 505 – 518

841 Tatsumi, Y. 2005. The subduction factory: how it operates in the evolving earth. GSA

842 today. 15 (7): 4 – 10.

843 Tatsumi, Y., Stern, R.J. 2006. Manufacturing continental crust in the subduction

844 factory. Oceanography. 19 (4): 104 – 112.

845 Tatsumi, Y., Takahashi, T. 2006. Operation of the subduction factory and

846 production of andesite. Journal of Mineralogical and Petrographical Sciences. 101:

847 145 – 153.

848 Thorpe, R.S. 1984. The tectonic setting of active Andean volcanism. In Andean

849 magmatism: Chemical and Isotopic Constraints (Harmon, R.S.; Barreiro, B.A.;

850 editors). Shiva Geological Series, Shiva Publications, Nantwich, U.K. 4 – 8.

36

851 Thorpe, R.S., Francis, P.W. 1979. Variations in Andean andesite composition and

852 their petrogenetic significance. Tectonophysics. 57: 53 – 70.

853 Thorpe, R. S., Francis, P. W, Hammill, M., Baker, M.C.W., 1982. The Andes.

854 Andesites. Ed Thorpe, R.S. 187 – 205.

855 Toro, G; Restrepo, JJ; Poupeau, G; Saenz, E y Azdimousa, A. 1999. Datación por

856 trazas de fisión de circones rosados asociados a la secuencia volcano –

857 sedimentaria de Irra (Caldas). Boletín de Ciencias de la Tierra. 13: 28 – 34.

858 Trenkamp, R., Kellogg, J.N., Freymueller, J.T., Mora, H.P., 2002. Wide plate margin

859 deformation, southern Central America and northwestern South America, CASA

860 GPS observations. Journal of South American Earth Sciences. 15: 157 – 171.

861 Uribe – Mogollón, C.A., 2013. Hydrothermal evolution of the Titiribí mining district.

862 Undergraduate thesis. EAFIT University. 127 pp.

863 van der Hammen, T., 1960. Estratigrafía del Terciario y Maastrichtiano continentales

864 y tectogénesis de los Andes Colombianos, Informe No. 1279. Servicio Geológico

865 Nacional, Bogotá. 128 pp.

866 Vargas, C., Mann, P.2013. Tearing and breaking off of subducted slabs as the result

867 of collision of the Panama arc-indenter with northwestern South America. Bull.

868 Seismol. Soc. Am. 103:2025–2046.

869 Vermeesch, P. 2009. RadialPlotter: A Java application for fission track,

870 luminescence and other radial plots. Radiation Measurements. 44: 409 – 410.

37

871 Walker, G.P.L., Wilson, C.J.N., Froggat, P.C. 1991. An ignimbrite veneer deposits;

872 the trail marker of a pyroclastic flow. Journal of Volcanology and Geothermal

873 Research. 9: 409 – 421.

874 Weber, M.B.I., Tarney, J., Kempton, P.D. and Kent, R. W., 2002, Crustal make-up of

875 the northern Andes: evidence based on deep crustal xenolith suites, Mercaderes,

876 SW Colombia. Tectonophysics. 345: 49 – 82.

877 White, W.M. 2013. Geochemistry (1st ed). Hoboke, NJ, Wiley-Blackwell. 660 pp.

878 White, W.M., Dupre, B., Vidal, P. 1985. Isotope and trace element geochemistry of

879 sediments from the Barbados Ridge – Demerara Plain region, Atlantic Ocean.

880 Geochim. Cosmochim. Acta. 49: 1875 – 1886.

881 Wilson, M. 1989. Igneous petrogenesis: a global tectonic approach. London, UK,

882 Chapman & Hall. 466 pp.

883 Winter, J.D. 2001. Introduction to igneous and metamorphic petrology. Upper Saddle

884 River, New Jersey, Prentice Hall. 796 pp.

885 Wörner, G.; Davidson, J.; Moorbath, S.; Turner, T.L.; McMillan, N.; Nye, C.; López-

886 Escobar, L.; Moreno, H. 1988. The Nevados de Payachata Volcanic Region

887 18°S/69°W, Northern Chile. I. Geological, geochemical and isotopic observations.

888 Bulletin of Volcanology. 30: 287 – 303.

889

890

891 Figure captions

892 Fig. 1 Overview map of the Colombia Andes, showing the subduction of the Nazca 893 and Caribbean plates beneath the South American plate. Also shown are the

38

894 Western, Central and Eastern Cordilleras and areas of Pliocene to present volcanic 895 activity, as well as the location of the study area (map from GeoMapApp, 896 http://www.geomapapp.org/).

897 Fig. 2 Geological map of the Amagá basin with the locations of the Cerro Amarillo 898 (CA), Anzá-Bolombolo (AB) and La Metida Creek (MC) sections (modified after 899 Sierra and Marín-Cerón, 2011).

900 Fig. 3 Cerro Amarillo stratigraphic of the eastern Amagá basin.

901 Fig. 4 Anzá Bolombolo stratigraphic section of the western Amagá basin.

902 Fig. 5 La Mertida Creek stratigraphic section of the western Amagá basin.

903 Fig. 6 Combined fission-track data radial plots of A) apatite fission-track data and B) 904 zircon fission track data with central ages.

905 Fig. 7 A) Streckeisen (1976) volcanic rock classification diagram. Both the Anzá- 906 Bolombolo and Cerro Amarillo samples plot in the andesite, basalt field. B) 907 Pyroclastic rock classification diagram of Pettijohn (1975), in which almost all La 908 Metida Creek and Anzó-Bolombolo pyroclastic rock samples plot in the crystal tuff 909 field.

910 Fig. 8 A) AFM diagram plot (Irvine and Baragar, 1971) for volcanic rocks from the 911 Cerro Amarillo and Anzá- Bolombolo sections. For comparison, data of Marriner and 912 Millward (1984), Ordoñez (2001) and Leal-Mejía (2011) from the Combia Formation 913 are also shown. B) Total alkalis versus silica (TAS) diagram of volcanic rocks from 914 the Cerro Amarillo and Anzá Bolombolo sections; boundaries in the total alkalis are 915 from LeMaitre et al. (1989) for rock classification and Irvine and Baragar (1971) for 916 magma series classification (red solid line). For comparison, data of Marriner and 917 Millward (1984), Ordoñez (2001) and Leal-Mejía (2011) are shown.

918 Fig. 9 A) Primordial mantle (McDonough and Sun, 1995) normalized trace element 919 spider diagrams of volcanic rocks from the Cerro Amarillo and Anzá-Bolombolo 920 sections. B) Chondrite (Evensen et al., 1978) normalized Rare Earth Element (REE) 921 patterns of the volcanic rocks from the Cerro Amarillo section. C) For comparison 922 trace element patterns of published data by Leall-Mejia (2011) from Combia 923 Formation and hypabyssal porphyritic intrusions in the study area, as summarized in

39

924 Marín-Cerón et al. (2019). Trace element data normalized after Wood et al. (1979). 925 D) For comparison, published REE data of Leal-Mejía (2011) normalized after Sun 926 and McDonough (1989) for hypabyssal porphyritic intrusions, Combia Formation 927 volcanic rocks and recent to present volcanism in the Central Cordillera, as 928 summarized in Marín-Cerón et al. (2019). E) The La/Sm versus Ba/Th plot indicates 929 that the Cerro Amarillo volcanic rocks were more derived from magma related to slab 930 dehydration, whereas the Anzá-Bolombolo samples trend more towards sediment 931 melting derived magma.

932 Fig. 10 Comparison between trace elements and isotopic systematics: A) Sr/Th 933 versus 87Sr/86Sr, and B) Ba/Th versus 143Nd/144Nd, for volcanic rocks of the CA 934 section (blue circles) and the AB section (red circles). The dashed line represents 935 altered oceanic crust values (AOC) (after Barret, 1983). Both diagrams show that 936 addition of fluids to the magma affected the geochemistry to the analyzed rocks and 937 that the slab dehydration effects are more pronounced in the eastern than the 938 western Amagá basin. C) Nd vs. Sr isotope ratio plot, showing the fields for the 939 Northern Volcanic Zone (NVZ), Central Volcanic Zone (CVZ) and Southern Volcanic 940 Zone (SVZ) as well as typical MORB composition, based on data from James et al. 941 (1976); Hawkesworth et al. (1979); James (1982); Harmon et al. (1984); Frey et al. 942 (1984); Thorpe (1984); Hickey et al. (1986); Hildreth and Moorbath (1988); Wörner et 943 al. (1988); Walker et al. (1991); de Silva (1991); Kay et al. (1991); Davidson and de 944 Silva (1992). Winter (2001); Marín-Cerón (2007), as summarized by and plot 945 modified from Marín-Cerón et al. (2019). The data of our study are shown for the 946 Cerro Amarillo section (yellow circles) and the Anzá-Bolombolo section (red circles), 947 plotted over the NVZ field.

948 Fig. 11 Lead isotopic systematics of the Combia Formation shown for the Cerro 949 Amarillo section (yellow circles) and the Anzá-Bolombolo section (red circles). A) 950 Plots of 208Pb/204Pb vs 206Pb/204Pb for the Andean volcanic zones (Northern Volcanic 951 Zone – NVZ; Central Volcanic Zone – CVZ; and Southern Volcanic Zone – SVZ) and 952 the pre-Andean basement (plot modified from data compilation plot of Marín-Cerón 953 (2019). Pacific sediments (Dasch, 1981; White et al., 1985); Paleozoic basement 954 (Chiaradia and Fontboté, 2002); metalliferous sediments from DSDP leg 92 (Barret 955 et al., 1987). B) Zoom on the 208Pb/204Pb vs 206Pb/204Pb diagram. Squares represent 956 possible end-members and the respective trends of the interaction between each of

40

957 the components involved during magma formation. AOC: altered oceanic crust; HS: 958 hemipelagic sediments; CS: carbonaceous sediments; LCC: lower continental crust. 959 The solid blue line represents the linear trend of the samples suggesting bimodal 960 mixing between primary magma (yellow star) and LCC. This plot is based on the 961 compilation of Marín-Cerón et al. (2019); Cretaceous Domain (Kerr, 2003); Lower 962 crust xenotiths (Weber et al., 2002); and ACC from Hole 504 (Pedersen and Furnes, 963 2001); NVZ data from Marín-Cerón (2007). The data from our study study are shown 964 for the Cerro Amarillo section (yellow circles) and the Anzá-Bolombolo section (red 965 circles) and the plotted over the Northern Volcanic Zone field.

966 Fig. 12 Schematic diagram for magma source genesis of Combia Formation 967 volcanism. AOC: altered oceanic crust; HS: hemipelagic sediments; CS: 968 carbonaceous sediments; AFC: assimilation fraction crystallization processes 969 (DePaolo, 1981); MASH: melting, assimilation, storage, and homogenization 970 processes (Hildreth & Moorbath, 1988); LCC: lower continental crust; UCC: upper 971 continental crust. Illustration based on model proposed for southwestern Colombian 972 volcanism (Marín-Cerón, 2007) and is compatible with models presented in Marín- 973 Cerón et al. (2019).

974

975

976 Tables

977 Table 1 Apatite fission-track data of the Combia Formation

978 Table 2 Zircon fission-track data of the La Metida Creek Formation

979 Table 3 Petrographic modal analyses

980 Table 4 Major elements (wt%) of the Cerro Amarillo and Anzá-Bolombolo section 981 samples

982 Table 5 Trace elements (ppm) of volcanic rock samples of the Cerro Amarillo and 983 Anzá-Bolombolo sections

984 Table 6 Isotopic compositions of volcanic rocks from the Cerro Amarillo and Anzá- 985 Bolombolo sections.

41

986 Table 7 Estimated isotopic compositions and parameters for end-member involved in 987 magma genesis

42

Table 1 Apatite fission‐track data of the Combia Formation

Number Single grain Age Central Sample RhoS RhoI RhoD U (ppm) Lithology of age range Ns Ni P(χ2) (%) dispersion age* (Ma) number (x105t/cm2) (x105t/cm2) (x105t/cm2) ±2 SE grains (Ma) (%) ±2 SE AB section

Pyroclastic JJ13 60 2.8 – 36.9 0.28 67 4.88 1183 9.72 44.9 2.7 7.9±2.1 6±0 agglomerate

JJ22 Tuff 35 3.2 – 38.5 0.29 35 4.96 592 9.72 36.3 24.0 8.4±3.1 7±1

MC section

JJ17 Lapilli – tuff 60 3.6 – 71.5 0.26 36 4.57 624 9.72 7.3 50.8 8.3±3.1 6±0

JJ14 Tuff 33 4.9 – 374.9 0.29 18 3.34 209 9.72 0.0 141.1 15.9±11.1 4±1

JJ21 Tuff 51 3.5 – 74.2 0.17 19 4.56 517 9.72 66.9 2.0 5.1±2.5 6±1

Volcaniclastic JJ18 16 4.7 – 48.9 0.23 7 4.59 139 9.72 0.9 155.4 7.3± 9.4 6±1 sandstone

JJ19 Tuff 20 6.6 – 138.7 0.24 10 3.23 133 9.72 45.7 2.1 10.5±7.3 4±1

JJ20 Lapilli tuff breccia 27 4.9 – 138.5 0.50 28 5.68 317 9.72 6.9 57.1 15.8±9.2 9±1

combined 302 2.8‐374.9 9.72 0.0 48.6 8.6±1.4 6±0

Note – RhoS: spontaneous track density. RhoI: induced track density; P(χ2): Chi2 probability. Fission‐track ages were calculated using a Zeta value of 288.36±7.90. AFT data calculated with Binomfit of Brandon (see Ehlers et al., 2005).

Table 2 Zircon fission‐track data of the La Metida Creek Formation

Number Single grain Age Central Sample RhoS RhoI RhoD P(χ2) U (ppm) Lithology of age range Ns Ni dispersion age* (Ma) number (x105t/cm2) (x105t/cm2) (x105t/cm2) (%) ±2 SE grains (Ma) (%) ±2 SE Lapilli tuff JJ1 30 1.0 – 14.3 4.65 175 17.6 662 2.71 24.4 24.2 6.7±1.5 258±21 breccia

JJ8 Tuff 12 2.1 – 19.4 5.89 58 24.1 237 2.67 3.3 44.3 6.2±2.6 359±47

JJ3 Tuff 25 4.0 – 19.5 10.6 383 27.5 993 2.71 47.5 6.6 10.0±1.7 405±28

JJ9 Tuff 28 0.9 – 25.5 13.2 570 33.1 1432 2.67 0.0 35.3 10.1±2.1 493±29

JJ10 Tuff 22 5.0 – 20.0 10.5 363 28.6 989 2.66 68.0 0.6 9.3±1.5 427±29

JJ4 Tuff 31 2.9 – 25.8 17.7 585 38.8 1280 2.70 0.3 24.1 11.2±2.0 572±36

Volcaniclastic JJ6 51 2.9 – 46.7 10.6 909 22.1 1844 2.69 0.0 44.1 12.7±2.4 327±17 sandstone

JJ5 Tuff 31 2.6 – 14.8 14.1 515 3.85 1405 2.69 73.1 6.5 9.4±1.4 569±14

Lapilli tuff JJ11 48 1.3 – 16.1 6.27 362 2.63 1517 2.65 14.2 19.6 6.1±1.1 394±22 breccia

Lapilli tuff JJ7 68 4.3 – 14.8 9.00 824 29.5 2705 2.68 94.0 0.7 7.8±1.1 439±20 breccia

Combined 346 0.9 – 46.7 10.2 4744 28.5 13206 2.68 0.0 33.9 9.1±1.1 418±14

Note – RhoS: spontaneous track density. RhoI: induced track density; P(χ2): Chi2 probability. Fission‐track ages were calculated using a Zeta value of 191.6±10.25. ZFT data calculated with Binomfit of Brandon (see Ehlers et al., 2005).

Table 3 Petrographic modal analyses CA section CA - 1 CA – 2 CA – 7 CA – 13 CA - 14 CA - 18 CA - 22 Plagioclase 48.61 19.1 36.2 9.25 32.1 89.7 31.2 Hypersthene --- 42.2 23.5 51.2 47.2 10.3 21.8 Augite 13.19 4.7 3.2 9 9.89 --- 9.5 Hornblende 3.47 0.6 ------5.4 Calcite 9.03 1.3 ------7.58 --- 5.01 Olivine 5.21 ------8 Biotite --- 0.5 0.32 1 3.23 --- 1.6 Sericite ------3.5 ------Rock fragments 19.8 19.4 28.6 29.55 ------17.49 Oxides --- 8.8 4.8 ------Opaque minerals 0.69 3.4 ------Total 100 100 100.12 100 100 100 100 AB section AB – 2 AB – 6 AB – 7 Plagioclase 57.14 69.13 52.22 Pyroxene 42.86 12.01 29.1 Spherulites --- 9.86 18.68 Olivine --- 9 --- Total 100 100 100 MC section QML - 1 QML – 6 QML - 7 QML – 9 QML – 10 QML - 14 QML - 18 QML - 22 QML - 28 QML - 31 Plagioclase 71.22 61.4 55 40 57.14 85 57.5 62.5 60 77.14 Pyroxene --- 10.53 15 20 22.85 7.6 25 25 25 14.29 Hornblende 5.91 5.26 5 ------2.4 ------Biotite 0.33 ------Oxides 2.13 ------5.71 5.72 5 --- 12.5 15 8.57 Rock fragments 16.48 15.79 --- 8.57 ------Opaque minerals 3.94 7.02 25 25.72 14.29 ------Spherulites ------17.5 ------Total 100.01 100 100 100 100 100 100 100 100 100

Table 4 Major elements (wt %) of the Cerro Amarillo and Anzó – Bolombolo section samples

Oxides JJ2‐1‐ JJ2‐1‐ JJ1 ‐ 3 JJ1 ‐ 6 JJ1 – 9 JJ1 ‐ 13 JJ1 ‐ 17 JJ1 – 18 JJ1 ‐ 20 JJ1 ‐ 23 JJ3 – 2 JJ3 ‐ 5 JJ3 ‐ 6 JJ3 ‐ 9 JJ4 ‐ 2 JJ4 ‐ 3 (wt %) 10 14

SiO2 53.20 52.68 52.56 52.62 52.55 52.10 52.49 52.64 52.09 51.19 52.07 52.22 51.78 52.34 55.45 55.78

Al2O3 14.80 14.70 14.82 14.71 14.66 14.61 14.59 14.87 18.30 15.29 15.05 15.22 17.49 17.66 19.02 18.93

Fe2O3 t 12.03 11.77 12.13 12.82 12.67 12.55 12.93 12.75 10.40 13.44 13.48 13.63 10.49 10.23 5.01 5.53 MnO 0.19 0.17 0.17 0.19 0.19 0.18 0.19 0.18 0.18 0.21 0.21 0.21 0.17 0.17 0.09 0.09 MgO 3.56 3.51 3.32 3.48 3.43 3.32 3.49 3.30 2.99 3.44 3.50 3.46 3.48 3.05 1.05 1.12 CaO 7.68 7.82 7.56 7.57 7.58 7.47 7.36 7.50 9.33 8.28 8.35 8.24 8.80 8.84 4.51 5.28

Na2O 2.95 2.74 2.63 2.60 2.69 2.70 2.76 2.80 2.66 2.76 2.89 2.80 2.94 2.98 3.58 3.73

K2O 1.89 1.75 1.94 1.88 1.98 1.84 1.79 1.82 1.20 1.40 1.25 1.45 1.47 1.48 3.58 4.60

TiO2 1.31 1.30 1.38 1.39 1.38 1.38 1.38 1.38 0.86 1.03 1.06 1.06 0.96 0.99 5.10 0.52 P2O5 0.48 0.48 0.50 0.50 0.51 0.52 0.51 0.28 0.39 0.39 0.39 0.44 0.46 0.47 0.49 LOI 1.48 2.75 2.60 2.19 1.77 1.98 1.47 1.74 1.12 1.63 1.63 1.65 2.08 2.31 5.18 2.70 SUM 99.67 99.11 99.95 99.4 98.64 98.97 99.49 99.41 99.06 99.88 100.33 100.1 100.51 98.45 98.77 99.67

Table 5 Trace elements (ppm) of volcanic rock samples of the Cerro Amarillo and Anzó Bolombolo sections.

JJ1‐3 JJ1‐6 JJ1‐9 JJ1‐13 JJ1‐17 JJ1‐18 JJ1‐20 JJ1‐23 JJ3‐2 JJ3‐5 JJ3‐6 JJ3‐9 JJ4‐2 JJ4‐3 JJ2‐1‐10 J22‐1‐14

Li 8.52 8.33 7.58 8.46 8.73 9.75 8.71 10.4 8.98 8.76 7.64 9.39 8.07 8.38 13.6 15.5

Sc 31 30.7 30.2 30.5 30.1 29.5 30.1 30.4 27.4 33 32.8 33.1 24 24.8 8 8.75

Ti 7840 7760 8270 8340 8210 8110 8250 8240 5070 6170 6160 6200 5690 5900 2710 3020

V 343 340 367 369 361 365 366 367 296 331 338 337 268 278 96.9 100

Cr 12.7 11.6 13.5 13.1 13.2 13.5 13.5 13.3 10.9 15.7 16.5 17.2 31.4 32.6 1.4 1.02

Co 33.1 30.7 33.7 34.5 34.9 35.2 34.7 35.5 25.4 34.7 34.2 35.6 28 26.3 9.58 11.8

Ni 14.9 13 16 16.5 17.1 17.3 16.6 18 7.18 9.99 9.79 10.3 20.2 17.7 2.35 2.63

Cu 224 203 227 220 223 240 226 238 127 184 169 179 191 187 117 143

Zn 107 136 116 120 114 116 115 116 86.6 111 112 114 100 96.9 70.3 74.2

As 25.1 6.22 6.78 3.1 7.12 10.5 6.82 8.08 3.19 4.37 4.68 4.61 5.37 5.99 4.38 3.9

Rb 43.9 42.3 49.9 51.4 49.2 51.5 49.3 50.1 25.8 32.9 34.7 34 37.2 38.3 169 130

Sr 475 478 470 457 459 461 453 457 558 497 457 473 483 489 1870 1170

Y 26 26 28 27.5 27.2 27.6 27.3 27.4 18.9 23.3 23.5 24 21 22.2 15.5 17

Zr 110 108 116 115 114 116 114 115 55.7 75 76.2 77 65.7 68.5 106 118

Nb 5.97 5.83 6.18 6.25 6.11 6.26 6.19 6.16 2.63 3.03 3.07 3.06 3.05 3.16 4.86 5.54 JJ1‐3 JJ1‐6 JJ1‐9 JJ1‐13 JJ1‐17 JJ1‐18 JJ1‐20 JJ1‐23 JJ3‐2 JJ3‐5 JJ3‐6 JJ3‐9 JJ4‐2 JJ4‐3 JJ2‐1‐10 J22‐1‐14

Cd 0.0498 0.108 0.0574 0.0541 0.0487 0.0538 0.055 0.058 0.0441 0.0513 0.0469 0.0488 0.045 0.0461 0.0391 0.049

Cs 1.51 1.73 1.86 1.97 1.76 1.96 1.84 2 0.884 1.12 1.23 1.19 1.35 1.59 3 1.11

Ba 967 948 1030 990 1030 1010 1000 1010 607 842 823 847 717 742 1350 1250

La 10.5 10.5 11.2 11.2 11.1 11.3 11.2 11.2 5.7 7.46 7.54 7.62 7.14 7.39 19.7 21.4

Ce 22.7 22.6 23.9 24.3 23.8 24 23.9 23.9 12.4 16.2 16.4 16.5 15.4 15.7 36.5 39.9

Pr 3.21 3.21 3.42 3.46 3.36 3.42 3.41 3.41 1.83 2.39 2.41 2.42 2.22 2.29 4.69 5.13

Nd 14.7 14.6 15.5 15.8 15.5 15.7 15.4 15.4 8.64 11.3 11.5 11.7 10.3 10.7 18.4 20.2

Sm 3.87 3.94 4.24 4.29 4.16 4.2 4.25 4.18 2.48 3.24 3.33 3.37 2.81 2.92 3.74 4.13

Eu 1.21 1.23 1.27 1.27 1.26 1.28 1.28 1.24 0.858 1.09 1.07 1.1 0.98 1.04 1.2 1.29

Gd 4.53 4.46 4.74 4.71 4.73 4.67 4.66 4.73 2.96 3.87 3.92 3.96 3.34 3.52 3.3 3.59

Tb 0.713 0.717 0.747 0.758 0.745 0.754 0.755 0.753 0.479 0.623 0.645 0.66 0.542 0.58 0.456 0.487

Dy 4.5 4.55 4.65 4.77 4.68 4.85 4.78 4.78 3.19 4.08 4.18 4.23 3.58 3.77 2.54 2.84

Ho 0.925 0.929 1 1.01 0.984 0.986 0.975 0.984 0.665 0.844 0.869 0.868 0.746 0.789 0.503 0.548

Er 2.75 2.65 2.84 2.89 2.84 2.89 2.86 2.87 1.99 2.48 2.53 2.57 2.23 2.32 1.44 1.62

Tm

Yb 2.53 2.55 2.64 2.71 2.64 2.69 2.65 2.67 1.87 2.32 2.36 2.4 2.13 2.18 1.39 1.59 JJ1‐3 JJ1‐6 JJ1‐9 JJ1‐13 JJ1‐17 JJ1‐18 JJ1‐20 JJ1‐23 JJ3‐2 JJ3‐5 JJ3‐6 JJ3‐9 JJ4‐2 JJ4‐3 JJ2‐1‐10 J22‐1‐14

Lu 0.375 0.368 0.397 0.403 0.394 0.398 0.394 0.395 0.271 0.331 0.347 0.349 0.313 0.329 0.211 0.239

Hf 2.99 2.92 3.17 3.18 3.15 3.15 3.1 3.12 1.61 2.18 2.25 2.27 1.81 1.87 2.55 2.81

Ta 0.368 0.37 0.39 0.39 0.388 0.392 0.385 0.384 0.166 0.191 0.189 0.194 0.192 0.193 0.282 0.303

Tl 0.298 0.234 0.322 0.36 0.361 0.297 0.245 0.272 0.2 0.268 0.317 0.222 0.282 0.297 0.351 0.138

Pb 9.64 9.7 10.2 10.5 10.1 10.2 10.2 10.1 6 9.58 9.9 9.67 4.63 4.74 10.9 12.1

Th 1.98 1.97 2.07 2.11 2.06 2.1 2.08 2.09 1.03 1.39 1.4 1.41 1.19 1.23 4.46 4.95

U 1.08 1.08 1.18 1.11 1.12 1.11 1.1 1.11 0.517 0.854 0.857 0.843 0.528 0.532 1.22 2.19

Table 6 Isotopic compositions of volcanic rocks from the Cerro Amarillo and Anzó Bolombolo sections.

JJ1‐3 JJ1‐9 JJ1‐17 JJ1‐23 JJ3‐2 JJ3‐5 JJ3‐9 JJ4‐2 JJ2‐1‐10 JJ2‐1‐14

87Sr/86Sr 0.703911 0.703930 0.703931 0.703921 0.703916 0.703876 0.70387 0.703862 0.70417 0.70416

2σ 0.000006 0.000008 0.000008 0.000008 0.000006 0.000006 0.000006 0.000006 0.000006 0.000006

87Sr/86Sr 0.70391 0.70393 0.70393 0.70393 0.70392 0.70388 0.70387 0.70387 0.70417 0.70417 (leached)

2σ 0.000006 0.000006 0.000006 0.000008 0.000006 0.000006 0.000006 0.000006 0.000006 0.000006

143Nd/144Nd 0.51294 0.51293 0.51292 0.51295 0.51293 0.51296 0.51293 0.51298 0.51290 0.51289

2σ 0.000009 0.000018 0.000017 0.000010 0.000010 0.000007 0.000013 0.000008 0.000007 0.000010

εNd 5.959 5.930 5.684 6.223 5.846 6.420 5.905 6.882 5.255 5.191

208Pb/204Pb 38.72422 38.73927 38.73610 38.73702 38.71385 38.67951 38.68273 38.79674 38.74386 38.75921

2σ 0.00418 0.00418 0.00418 0.00418 0.00418 0.00417 0.00417 0.00418 0.004179 0.004180

207Pb/204Pb 15.61989 15.62459 15.62410 15.62365 15.62449 15.61989 15.62098 15.61286 15.60328 15.61037

2σ 0.00167 0.00167 0.00167 0.00167 0.00167 0.00167 0.00167 0.00167 0.001668 0.001669

206Pb/204Pb 18.96667 18.97333 18.97308 18.97264 18.93568 18.91002 18.91054 19.06522 19.00789 19.018005

2σ 0.00365 0.00365 0.00365 0.00365 0.00365 0.00364 0.00364 0.00367 0.003662 0.003666

Table 7 Estimated isotopic compositions and parameters for end‐member involved in magma genesis 208Pb/204Pb 206Pb/204Pb Pb (ppm) 142Nd/144Nd Nd (ppm) 87Sr/86Sr Sr (ppm)

AOC1,9 38.14 18.59 0.53 0.51280 4.71 0.70381 61.40

HS2,3,4 38.86 18.64 9.59 0.51247 17.0 0.70763 336.16

CS2,3,4,5 38.16 18.46 3.70 0.51242 0.89 0.70858 1 504.12

SC_1* 38.62 18.62 1.43 0.51285 5.94 0.70454 75.138

SC_2* 38.57 18.60 1.54 0.51284 1.15 0.70661 146.412

MW6,7 37.90 18.50 0.02 0.51310 0.71 0.70270 9.80

PM* 38.45 18.59 0.82 0.51307 5.81 0.70469 114.95

LCC8 38.73 19.02 8.81 0.51309 9.50 0.70423 173.00

* Estimated values. Kd from Halliday et al. (1995). AOC: Altered oceanic crust; HS: Hemipelagic sediments; CS: Carbonaceous sediments; SC_1: Subduction component 1 (AOC+HS); SC_2: Subduction component 2 (SC_1 + CS); MW: Mantle wedge (5%cpx, 25%opx, 70%ol); PM: Primary magma; and LCC: Lower continental crust. Values are calculated based on procedures followed by Marín‐Cerón (2007) for the SW Colombian volcanic arc. Values for AOC, HS, CS, MW, and LCC, are given according to data presented by Marín‐Cerón (2007). SC_2 represents the interaction between AOC (85%) and sediments (HS – 10% and CS – 5%). PM is calculated assuming 40% of metasomatized mantle by subduction component. It is considered 10% of LCC interacts with primary magma for magma source genesis. 1Pedersen et al. (2001); 2Plank and Langmuir (2000); 3Patiño et al. (2000); 4Vervoort et al. (1999); 5Hemming and McLennan (2001); 6Salters and Stracke (2004); 7Saunders et al. (1988); 8Weber et al. (2002); 9 10 Barret (1983); and Halliday et al. (1995).

12°N Caribbean plate

11°N

10°N

9°N Panama Venezuela

8°N Colombia

7°N Medellín AB South Nazca plate MVB Study American 6°N 6 cm/yr area Fig. 2 plate EC Paipa-Iza Nevado del Ruiz 5°N WC CPB Nevado Bogotá del Tolima EC = Eastern Cordillera CC = Central Cordillera Cerro Machin 4°N Nevado WC = Western Cordillera de Huila AB = Antioquia batholith CC CPB = Cauca-Patia basin 3°N MVB = Magdalena Valley Azufral Cumbal basin 81°W 80°W 79°W 78°W 77°W 76°W 75°W 74°W 73°W 72°W 71°W 70°W 69°W

Fig. 1 Bernet et al. Paleozoic Mesozoic Cenozoic Lithological units Jurassic CretaceousOligocene-Miocene Quaternary CA =CerroCA Amarillo MC =LaMetdiaCreek AB =Anzá andBolombolo Pgnp Kida Kdhb Kdha Kcdu Ksga Png Peni Pbsd Ksta Kgh Tmc Tdsa Tada Tadh PEa Pei Pes Pev TRa Jdp Jgr Jus Kaa Kuh Ksc Kvc Kvb Kg Tdc Toi Tos Tsc Qt Qtl Qd Qar Td Caldas amphibolite Migmatite La Iguanamicaceous gneiss Ayurá Gr. Montebello interbedded Ayurá Gr. Montebello phyllite Ayurá Gr. Montebello green schists Ayurá Gr. Montebello clastictexture Palmitas gneissic granite Amagá stock Pueblito diorite Roneral gabbro Ultramafic rocks Altavista batholith Penderisco Fm Ultramafic harzburgite Quebradagrande Fm. memeber sedimentary Quebradagrande Fm. volcanic memeber Barroso Fm. Heliconia hornblendediorite Heliconia quartzdiorite Antioquia batholith Altamira gabbro gabbroHispania Ursula stock Camburmbia stock Lower AmagáFm. Upper AmagáFm. Combia Fm. volcanic member Combia Fm. volcaniclastic member Andesitic dikesandsills porphyry Augite-andesitic Andesitic porphyry Dacitic porphyry Depris deposits Recent alluvialdeposits Gabbro Terrace deposits Talus deposits

Ursula stock 5°47’30’’N 5°58’30’’N 6°9’30’’N 6°15’0’’N 6°20’30’’N 6°4’0’’N 55’’N7°63’N7°10’ 75°34’30’’N 75°41’0’’N 75°46’30’’N 75°52’0’’N MC 55’’N7°63’N7°10’ 75°34’30’’N 75°41’0’’N 75°46’30’’N 75°52’0’’N AB 06 km Fig. 2Bernetetal. CA

5°53’0’’N 6°4’0’’N 6°9’30’’N 6°15’0’’N 6°20’30’’N 5°58’30’’N 5°47’30’’N

CERRO AMARILLO SECTION

c

i

s y

h Texture LEGEND

s

g

p

e

o

a

l

n r

i o

s k

l g

l

i

(m)

unit

h k c

i t

h

i

t i c

p a

s

h

r L o

a

l t

A

T L Petrography

B S Geochemistry Pyroclastic Lapilli tuff 200 breccia breccia

Scoria Ignimbrite

150 Vesicular Basaltic basaltic andesite andesite

Covered with vegetation

JJ001-23 CA-22 JJ001-20 100 JJ001-18 JJ001-17 JJ001-13 CA-18 JJ001-9

JJ001-6 CA-14

JJ001-3 CA-13 COMBIA FORMATION COMBIA

50

JJ003-2 CA-7

JJ003-5 JJ003-6

JJ003-9 CA-2

JJ004-2 30 m 30 CA-1 JJ004-3

Fig. 3 Bernet et al. ANZA - BOLOMBOLO SECTION

Trachy-andesite and pyroclastic sequence Tuffaceous sequence

c

c

i

i

s

h

h

y

s

y

p

p

g

e

g

t

a

a o

Texture i Texture

n

o

r

l

r

l

n

k

g

o

g

o i i

i

(m)

unit i

s

s

c

AFT AFT u

l

l

t

(m)

t i h

l

l h

k k

i

t i

t

a

i

a h i

c

c

r analyses

p

p

r

t

L

t T L

o o

a

a Thickness

l

l Petrography Ash Petrography

Ash

S S

L L

B

B Geochemistry

10 Unconformity 10

8 8 JJ13

AB7

A

A I

I 6 6

B

B M

M Unconformity

O

O

CFORMATION CFORMATION

4 4

2 2 AB2 JJ2-001-10 AB6 JJ2-001-14

JJ22 1 m 1

LEGEND LITHOLOGY

Pyroclastic breccia Tuff Lapilli tuff and agglomerate Fig. 4 Bernet et al.

Pyroclastic Undiferentiated Trachy- flow lava flow andesite CONTACTS

Abrupt Erosional

Gradational LA METIDA CREEK SECTION

LEGEND

c

i

s s

h LITHOLOGY

s

e

p

r

e

a y

Texture u

n

r

t

g

k

g

c i

i

(m)

o

AFT unit

s

ZFT c

l

l

t

i u

l

k

i

r

o a

h Tuff

h Lapilli tuff

t analyses

c

r analyses

p

s

h

t

T

t S

o

a

Petrography i

l

A S

L L

B

QLM -1 Lapilli tuff Pyroclastic breccia flow QLM -6 QLM -7 Tuffaceous Volcaniclastic sandstone sandstone STRUCTURES QLM -9 Ripples Cross- Load structures 40 bedding QLM -10 Lenses Nodules Bio- turbation

Organic Fossils matter

JJ17 QLM -10

30 JJ1

JJ8 QLM -18 JJ3 JJ14 JJ9 QLM -22 JJ10

20 COMBIA FORMATION COMBIA

JJ4 JJ21 QLM -28 JJ6 JJ18 JJ5 JJ19 10 JJ11

QLM -31

JJ7 JJ20 5 m 5

Fig. 5 Bernet et al. ŽŵďŝĂ&ŽƌŵĂƟŽŶĐŽŵďŝŶĞĚ&dĚĂƚĂ;ŶсϯϬϮͿ ŽŵďŝĂ&ŽƌŵĂƟŽŶĐŽŵďŝŶĞĚ&dĚĂƚĂ;ŶсϯϰϲͿ Central age = 8.54 ± 0.66 Ma (1 ʍͿ Central age = 9.1 ± 0.24 Ma (1ʍͿ 79Ma Dispersion = 34 % Dispersion = 52 % 319Ma 60 250 P(ʖϸͿсϬ͘ϬϬ 50 P(ʖϸͿсϬ͘ϬϬ 200 40 150 30 100 20

50 2 10 0 -2

2 0 -2

0.99Ma 0.7Ma

Ns+Ni 0 4 16 36 64 100 144 196 Ns+Ni 0 12 49 110 196 306 441

77.00U [ppm] 1590.00

Fig 6 Bernet et al. A) Quartz

Lava flows Cerro Amarillo Anzá – Bolombolo

Quartz andesite

Rhyolite Dacite

Alkali rhyolite Rhyodacite

Quartz Andesite Andesite, Quartz Quartz latite alkali Quartz latite quartz- basalt trachyte andesite/basalt trachyte basalt Alkali trachyte Trachyte Latite Latite andesite/basalt Alkali -feldspar Plagioclase

B) Glass

Pyroclastic rocks La Metida Creek section Anzá – Bolombolo section

Vitric tuff

Lithic tuff Crystal tuff

Rock fragments Crystals

Fig 7 Bernet et al. 12 - 6 Ma magmatism A) F F F Marriner & Millward (1984)

Ordoñez (2001)

A = Na2O + K2O Leal-Mejia (2011) Tholeiitic F = FeO Tholeiitic M = MgO Tholeiitic

Calc-alkaline

A M Calc-alkaline Calc-alkaline

Cerro Amarillo Anzá – Bolombolo A AM

B) 18 Ultrabasic Basic Intermediate Acid 12-6 Ma Marriner & Millward (1984) alkaline 12-6 Ma Ordoñez (2001) 17-9 Ma Leal-Mejia (2011) 15 24-20 Ma Leal-Mejia (2011)

Phonolite 12 P-N Trachyte

P-T Benmorite O (wt%)

2 9 Trachy Rhyolite Mugearite andesite B+T O+K Hawaiite 2 6 Nephelin Dacite Na

Andesite 3 Basalt B-A Anzá – Bolombolo Cerro Amarillo 0 35 45 55 65 75 SiO2 (wt%)

Fig. 8 Bernet et al. A) B)

1000 100 Trace elements Cerro Amarillo Rare Earth Elements Cerro Amarillo Anzá – Bolombolo Anzá – Bolombolo 100

10

10 Sample/REEchondrite Sample/Primitive mantle Sample/Primitive Combia Fm (this study) Combia Fm (this study) normalized to primordial mantle after McDonough and Sun (1995) normalized to chondrite after Evensen et al. (1978) 1 1 Cs RbBa Th U Nb Ta La Ce Pr PbNd Sr Sm Zr Hf Eu Gd Tb DyHo Y Er Li Yb Lu La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

C) 1000 D) 1000 17–6 Ma hypabyssal porphyritic rocks 17–6 Ma hypabyssal porphyriticrocks 12–6 Ma Combia Fm. 12–6 Ma Combia Fm. 100 3–0 Ma volcanicm 100

10

10

1 Sample/REEchondrite Sample/Primitive mantle Sample/Primitive Data from Leal-Mejia (2011) Data from Leal-Mejia (2011) normalized to chondrite after Sun and McDonough (1989) normalized to primordial mantle after Wood et al. (1979) 1

Cs Rb Ba Th UKNdTa Nb La CeSr P HfZr Sm Ti Tb Y La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

E) 1400

1200

1000 Slab

800 dehydration

Ba/Th 600

400 Cerro Amarillo Sediment 200 Anzá – Bolombolo melting 0 024681012 La/Sm

Fig. 9 Bernet et al. A) B) 1000 800 900 Addition of 700 effect of slab 800 aqueous dehydration 600 700 fluids to the magma eastern Amagá basin 600 500 500 400 Sr/Th 400 Ba/Th 300 western Amagá basin 300 Anzá – Bolombolo 200 200 Cerro Amarillo 100 Anzá – Bolombolo 100 Altered Oceanic Crust Cerro Amarillo 0 0 0.702 0.704 0.706 0.708 0.710 0.712 0.5118 0.5120 0.5122 0.5124 0.5126 0.5128 0.5130 0.5132 87Sr/86Sr 143Nd/144Nd C) MORB 0.5130 Northern Anzá – Bolombolo Volcanic Cerro Amarillo Zone 0.5128 Southern

Nd Volcanic

144 0.5126 Zone

Nd/ Compilation of typical

143 Andean volcanism 0.5124 isotopic compositions from Marin-Ceron et al. (2019)

0.5122 Central Volcanic Zone

0.5120 0.702 0.705 0.710 0.715 87Sr/86Sr

Fig 10 Bernet et al. 39.4 A) Paleozoic Precambrian basement 39.0 basement CVZ NVZ 38.6 CVZ Pb SVZ Galapagos Cretaceous 204 38.2 Pacific metaliferous and AOC basement Pb/ carbonate-rich sediments ODP hole & LCC 208 37.8 504B xenolith NVZ Present-day Andean volcanism 37.4 NVZ = Northern volcanic zone Anzá – Bolombolo CVZ = Central volcanic zone Cerro Amarillo SVZ = Southern volcanic zone 37.0 17.0 17.2 17.4 17.6 17.8 18.0 18.2 18.4 18.2 18.8 19.0 19.2 19.4 B) 206Pb/204Pb

39.0 Lower Hemipelagic Andean Continental sediments Paleozoic Crust basement 38.6

Pb ODP hole Paficic 504B

204 sediments Carbonaceous 38.2 Altered

Pb/ sediments oceanic crust 208

37.8 Mantle wedge Squares present possible end-members Anzá – Bolombolo Cerro Amarillo 37.4 18.2 18.4 18.6 18.8 19.0 19.2 19.4 206Pb/204Pb

Fig. 11 Bernet et al. Western Eastern Amagá basin Amagá basin boundary boundary

Anzá – Bolombolo Cerro Amarillo La Metida Creek Upper Continental Crust Lower Continental Crust

Mantle wedge Magmatic processes (AFC & MASH)

Subduction HS+CS Primary components AOC magma Sediment melting, dehydration/ decarbonation

Fig. 12 Bernet al.